Home
Patent Search
IMT Blog
REGISTER
|
SIGN IN
United States Patent Application
20020152267
Kind Code
A1
Lennon, Alison J.
October 17, 2002
Method for facilitating access to multimedia content
Abstract
A system is described in which a media browser (101), operating as a software application on a user terminal or preferably a server for a number of users, provides a user with a single user interface that facilitates browsing and searching different metadata collections over the Internet (102). A metadata server (212) is associated with each of the metadata collections. When the metadata server (212) receives a request from the media browser (101), the metadata server (212) interprets the request and replies with a description that satisfies the request and according to a predetermined scheme. The description contains at least one link which represents a return link which represents a return request to the metadata server (212). Specifically disclosed are methods (2000) for recording user location preferences at a service and using one or more lists of such preferences to focus searching responses to statistically desirable material. Monitoring the validity of a bookmarked location is also described.
Inventors:
Lennon; Alison J.
(Balmain, AU)
Correspondence Name and Address:
30 ROCKEFELLER PLAZA
FITZPATRICK CELLA HARPER & SCINTO
NEW YORK
NY
10112
US
Series Code:
023757
Filed:
December 21, 2001
U.S. Current Class:
709/203;
707/10
U.S. Class at Publication:
709/203;
707/10
Intern'l Class:
G06F 015/16;
G06F 017/30; G06F 007/00
Claims
We claim:
1. A method of facilitating access to multimedia content distributed over a computer network, said method comprising the steps of: at a server within said network, monitoring location recording activities of subscribers of said server to form a ranked list of multimedia locations recorded by said subscribers; receiving search criteria relating to said multimedia content entered by a subscriber; and applying said search criteria to said ranked list to provide said subscriber with network identifiers for said multimedia content corresponding to those said recorded locations best satisfying said search criteria.
2. A method according to claim 1 wherein said search criteria is applied to an entirety of said ranked list.
3. A method according to claim 1 wherein said search criteria is applied to a subset of the network identifiers comprising said ranked list.
4. A method according to claim 3 wherein said search set comprises a first predetermined number of highest ranked network identifiers from said list.
5. A method according to claim 4 further comprising the steps of randomly identifying from lower ranked ones of said network identifiers of said list a second predetermined number of said network identifiers, and substituting said second number of network identifiers into said search set, said second number being less than said first number.
6. A method according to claim 5 wherein said second number of network identifiers are randomly substituted into said search set.
7. A method according to claim 5 or 6 wherein said random selection of said lower ranked network identifiers is influenced by a subscriber selected parameter.
8. A method according to claim 7 wherein said parameter determines said second number of said network identifiers.
9. A method according to claim 8 wherein said random selection comprises, for each of said second numbers selecting a rank position from said list and selecting a lower ranked one of said network identifiers to be substituted into said search set.
10. A method according to any one of the preceding claims wherein said network identifier identifies an item of metadata, wherein said metadata represents one of a description of said item of multimedia content or a description of a collection of multimedia content.
11. A method according to any one of the preceding claims wherein said network identifiers represent descriptions of collections of multimedia content, in which individual said descriptions of said content are accessible via branch network identifiers depending from other said network identifiers.
12. A method according to claim 11 wherein said network identifiers are selected from the group consisting of a URI for an XML description, and a URI for a metadata server root.
13. A method according to any one of the preceding claims further comprising, at said server, examining a corresponding ranked list at another server within said network and modifying said ranked list with locations obtained from said corresponding ranked list.
14. A method according to claim I wherein said search criteria is generated at a browser coupled to said server.
15. A method of forming a searchable list of computer network locations, said method comprising the steps of: monitoring, at a server to a plurality of network subscribers, bookmarking activities of said subscribers with respect to certain ones of said network locations; forming a list of network identifiers bookmarked by individual ones of said subscribers; and ordering said network identifiers in said list according to a frequency of bookmarking by said subscribers.
16. A method according to claim 15 further comprising: monitoring accesses by individual ones of said subscribers to said network identifiers within said list to modify said frequency; and re-ordering said list according to said modified frequency.
17. A method of facilitating access to multimedia content distributed over a computer network, said method comprising, at a server within said network, the steps of: interfacing an application by which at least one subscribing user thereof is enabled to access said multimedia content, each itemn of said multimedia content being identified by a network identifier; monitoring individual said subscribing users' intent to use or store items of said multimedia content; and in response to said monitoring, forming a ranked list of network identifiers relating to said items for use in facilitating access to said multimedia content for a predetermined set of said subscribing users, wherein an ordering of said list is determined by a frequency that the corresponding said network identifier is referenced with intent to use or store by said subscribing users.
18. A method according to claim 17 wherein said predetermined set of said subscribing uscr comprises all said subscribing users.
19. A method according to claim 17 wherein said network identifier comprises a uniform resource identifier.
20. A method according to claim 17 wherein said network identifier comprises a uniform resource location.
21. A method according to claim 17 wherein each said item is identified directly by said network identifier.
22. A method according to claim 17 wherein each said item is referenced indirectly by said network identifier.
23. A method according to claim 17 comprising the further steps of: receiving a search request from one said application; identifying search criteria from said search request; applying said search criteria to metadata associated with network identifiers forming at least part of said ranked list; and transmitting search results to said application according to a correspondence of said results to the ranking of said list.
24. A method according to claim 23 wherein said applying and transmitting comprise at least one of: (1) applying said search criteria to said network identifiers in said list and transmitting said search results according to said ranking; (2) selecting a plurality of highest ranked locations of said list and applying said search criteria to said plurality; and (3) randomly mutating the ranked locations with locations lower in the rank list and applying the search criteria to the mutated locations.
25. A method according to claim 17 wherein said application is provided to a suer terminal by said server to thereby provide said interface.
26. A computer readable medium, having a program recorded thereon, where the program is configured to make a computer execute a procedure to facilitate access to multimedia content distributed over a computer network, said program comprising: code for monitoring location recording activities of subscribers of said server to form a ranked list of multimedia locations recorded by said subscribers; code for receiving search criteria relating to said multimedia content entered by a subscriber; and code for applying said search criteria to said ranked list to provide said subscriber with network identifiers for said multimedia content corresponding to those said recorded locations best satisfying said search criteria.
27. A computer readable medium according to claim 26 wherein said search criteria is applied to an entirety of said ranked list.
28. A computer readable medium according to claim 26 wherein said search criteria is applied to a subset of the network identifiers comprising said ranked list.
29. A computer readable medium according to claim 28 wherein said search set comprises a first predetermined number of highest ranked network identifiers from said list.
30. A computer readable medium according to claim 29 further comprising code for randomly identifying from lower ranked ones of said network identifiers of said list a second predetermined number of said network identifiers, and substituting said second number of network identifiers into said search set, said second number being less than said first number.
31. A computer readable medium according to claim 30 wherein said second number of network identifiers are randomly substituted into said search set.
32. A computer readable medium according to claim 30 or 31 wherein said random selection of said lower ranked network identifiers is influenced by a subscriber selected parameter that determines said second number of said network identifiers.
33. A computer readable medium according to claim 32 wherein said code for said random selection comprises, for each of said second number, code for selecting a rank position from said list and selecting a lower ranked one of said network identifiers to be substituted into said search set.
34. A computer readable medium according to claim 26 wherein said network identifier identifies an item of metadata, wherein said metadata represents one of a description of said item of multimedia content or a description of a collection of multimedia content and said network identifiers represent descriptions of collections of multimedia content, in which individual said descriptions of said content are accessible via branch network identifiers depending from other said network identifiers.
35. A computer readable medium according to claim 34 wherein said network identifiers are selected from the group consisting of a URI for an XML description, and a URI for a metadata server root.
36. A computer readable medium according to claim 26 further comprising, code for examining a corresponding ranked list at another server within said network and modifying said ranked list with locations obtained from said corresponding ranked list.
37. A computer readable medium, having a program recorded thereon, where the program is configured to make a computer execute a procedure to form a searchable list of computer network locations, said program comprising: code for monitoring, at a server to a plurality of network subscribers, bookmarking activities of said subscribers with respect to certain ones of said network locations; code for forming a list of network identifiers bookmarked by individual ones -of said subscribers; and code for ordering said network identifiers in said list according to a frequency of bookmarking by said subscribers.
38. A computer readable medium according to claim 37 further comprising: code for monitoring accesses by individual ones of said subscribers to said network identifiers within said list to modify said frequency; and code for reordering said list according to said modified frequency.
39. A computer readable medium, having a program recorded thereon, where the program is configured to make a computer execute a procedure to facilitate access to multimedia content distributed over a computer network, said program comprising, at a server within said network: code for interfacing an application by which at least one subscribing user thereof .is enabled to access said multimedia content, each item of said multimedia content being identified by a network identifier; code for monitoring individual said subscribing users' intent to use or store items of said multimedia content; and code for, in response to said monitoring, forming a ranked list of network identifiers relating to said items for use in facilitating access to said multimedia content for a predetermined set of said subscribing users, wherein an ordering of said list is determined by a frequency that the corresponding said network identifier is referenced with intent to use or store by said subscribing users.
40. A computer readable medium according to claim 39 wherein said predetermined set of said subscribing user comprises all said subscribing users.
41. A computer readable medium according to claim 40 wherein said network identifier comprises one of a uniform resource identifier, a uniform resource location.
42. A computer readable medium according to claim 39 wherein each said item is one of identified directly by said network identifier or each said item is referenced indirectly by said network identifier.
43. A computer readable medium according to claim 39 further comprising: code for receiving a search request from one said application; code for identifying search criteria from said search request; code for applying said search criteria to metadata associated with network identifiers forming at least part of said ranked list; and code for transmitting search results to said application according to a correspondence of said results to the ranking of said list.
44. A computer readable medium according to claim 43 wherein said code for applying and transmitting comprise at least one of- (1) code for applying said search criteria to said network identifiers in said list and transmitting said search results according to said ranking; (2) code for selecting a plurality of highest ranked locations of said list and applying said search criteria to said plurality; and (3) code for randomly mutating the ranked locations with locations lower in the rank list and applying the search criteria to the mutated locations.
45. A computer readable medium according to claim 39 wherein said application is provided to a user terminal by said server to thereby provide said interface.
46. A server computer system for facilitating access to multimedia content distributed over a computer network, said system comprising: means for monitoring location recording activities of subscribers of said server to form a ranked list of multimedia locations recorded by said subscribers; means for receiving search criteria relating to said multimedia content entered by a subscriber, and means for applying said search criteria to said ranked list to provide said subscriber with network identifiers for said multimedia content corresponding to those said recorded locations best satisfying said search criteria.
47. A server computer for forming a searchable list of computer network locations, said server computer comprising: means for monitoring, at a server to a plurality of network subscribers, bookmarking activities of said subscribers with respect to certain ones of said network locations; means for forming a list of network identifiers bookmarked by individual ones of said subscribers; and means for ordering said network identifiers in said list according to a frequency of bookmarking by said subscribers.
48. A server computer for facilitating access to multimedia content distributed over a computer network, said computer comprising: means for interfacing an application by which at least one subscribing user thereof is enabled to access said multimedia content, each item of said multimedia content being identified by a network identifier; means for monitoring individual said subscribing users' intent to use or store items of said multimedia content; and means for in response to said monitoring, forming a ranked list of network identifiers relating to said items for use in facilitating access to said multimedia content for a predetermined set of said subscribing users, wherein an ordering of said list is determined by a frequency that the corresponding said network identifier is referenced with intent to use or store by said subscribing users.
49. A browser application for operation upon a subscriber terminal, said application comprising means for signalling, to a server to which said application couples, a bookmarking of a location accessed by said application, said location being within a computer network incorporating said server.
50. A server operating within a computer network, said server having at least one user browser application associated therewith and via which said user browser application accesses locations within said network, said server comprising: means for receiving from said user browser application, bookmark information relating to a location recorded for subsequent access by said user browser application; means for integrating said bookmark information received from plural ones of said user browser applications to form a database of said bookmark information.
51. A server according to claim 50 further comprising: search engine means for searching said bookmark information.
52. A server according to claim 50 wherein said bookmark information comprises a location within said network of an source of multimedia content,
53. A server according to claim 52 wherein said bookmark information further comprises a count of a number of accesses to said location,
54. A method of facilitating access to multimedia content distributed over a computer network, said method comprising the steps of: at a server within said network, monitoring network location recording activities of subscribers of said server to form a ranked list of network identifiers recorded by said subscribers, said network identifiers being a source of multimedia content; receiving search criteria relating to said multimedia content entered by a subscriber; using said ranked list to determine a subset thereof; applying said received search criteria to said determined subset; and providing, to said subscriber of said server, results of applying said criteria to said determined subset, said results being provided according to an ordering of the corresponding originating network identifier in said list.
55. A method according to claim 54 wherein said network identifiers in said ranked list identify the location of descriptions of multimedia content.
56. A method according to claim 55 wherein said descriptions of multimedia content identify the network location of the corresponding multimedia content.
Description
FIELD OF THE INVENTION
[0001] The present invention relates generally to enabling access to electronically-accessible multimedia content and, in particular, to systems and processes by which access via browsing or searching to such content is facilitated through metadata of the content and operations associated with a user interface.
BACKGROUND
[0002] As network connectivity has continued its explosive growth, content providers are using the World Wide Web (the "Web") to provide access to their multimedia content (eg. images, video, audio, etc.). Unlike textual content, such as HTML pages, multimedia content is not directly accessible to standard Web search engines. These search engines can examine sites on the Web and extract information about the textual content of those sites. Such information is typically termed "metadata" which is data that describes or catalogues aspects of other data. The extracted information (metadata) can then provide users with access to that content using their customised metadata databases.
[0003] In the case of multimedia, content providers or distributors typically store information about the multimedia items to which they have access in metadata databases. The content providers then enable access to these databases by providing a search engine that users or customers can access from a Web site, typically the content provider/distributor's own Web site. Customers wanting to view, or maybe purchase, content that a content provider/distributor has access to, can visit the Web site and use the search engine to search the content provider/distributor's metadata database. Typically the metadata database contains visual identifiers of the content (eg. thumbnails, video abstracts, audio previews, etc) as part of the metadata. The user can then make decisions about which item(s) they may wish to purchase/use based on the metadata that is returned from their searches.
[0004] In many cases the multimedia content is digital and on-line, and potential customers can purchase the rights to use or purchase a copy of the desired multimedia item from the content provider/distributor's Web site. More often than not, this transaction is completed on the Web site and the potential customer can directly download their newly acquired content. However this model of providing access to multimedia content does not require that the content is on-line. For example, a potential customer might be able to purchase the rights to use, or a copy, of the desired content from the Web site but the content may be delivered to the potential customer by non-electronic means (ie. the postal system). Another variation is that the potential customer may be redirected from a distributor's site to the actual content provider in order to purchase and acquire a copy of the desired content. Other variations include the potential customer being directed to a physical location to purchase the content and being posted books containing the metadata associated with items to be purchased.
[0005] In all the abovementioned situations, the potential customer can only gain access to the content to which each content provider/distributor has access. If the potential customer wanted to perform a search across several different content providers/distributors, the potential customer would have to visit the Web site and use the search engine of each of the different content providers/distributors. Such actions are often time consuming and annoying because the potential customer must use a different search engine interface each time.
[0006] These problems have encouraged the development of very large metadata databases on the Web where a content distributor either purchases the rights to content or simply acts as a distributor for smaller content providers. Examples of such are the large image databases of Getty and Corbus. This approach has its own problems. Firstly, the approach does not scale because as the databases become very large, the search time increases. Further, typically all the metadata has to be structured in a similar fashion in order to contain the same metadata keys. However, such is not always desirable as different metadata may be more appropriate depending on the targeted use of the content. For example, images captured for geological purposes would require different metadata that those captured for holiday brochures. Thirdly, smaller content providers have no way to directly sell their content (ie. they are effectively forced to use the larger distributors).
[0007] It is an object of the present invention to ameliorate one or more disadvantages of prior art arrangements.
SUMMARY OF THE INVENTION
[0008] In accordance with one aspect of the present invention there is disclosed A method of facilitating access to multimedia content distributed over a computer network, said method comprising the steps of:
[0009] at a server within said network, monitoring location recording activities of subscribers of said server to form a ranked list of multimedia locations recorded by said subscribers;
[0010] receiving search criteria relating to said multimedia content entered by a subscriber; and
[0011] applying said search criteria to said ranked list to provide said subscriber with network identifiers for said multimedia content corresponding to those said recorded locations best satisfying said search criteria.
[0012] In accordance with another aspect of the present invention there is disclosed a method of forming a searchable list of computer network locations, said method comprising the steps of:
[0013] monitoring, at a server to a plurality of network subscribers, bookmarking activities of said subscribers with respect to certain ones of said network locations;
[0014] forming a list of network identifiers bookmarked by individual ones of said subscribers; and
[0015] ordering said network identifiers in said list according to a frequency of bookmarking by said subscribers.
[0016] In accordance with another aspect of the present invention there is disclosed a method of facilitating access to multimedia content distributed over a computer network, said method comprising, at a server within said network, the steps of:
[0017] interfacing an application by which at least one subscribing user thereof is enabled to access said multimedia content, each item of said multimedia content being identified by a network identifier;
[0018] monitoring individual said subscribing users' intent to use or store items of said multimedia content; and
[0019] in response to said monitoring, forming a ranked list of network identifiers relating to said items for use in facilitating access to said multimedia content for a predetermined set of said subscribing users, wherein an ordering of said list is determined by a frequency that the corresponding said network identifier is referenced with intent to use or store by said subscribing users.
[0020] In accordance with another aspect of the present invention there is disclosed a browser application for operation upon a subscriber terminal, said application comprising means for signalling, to a server to which said application couples, a bookmarking of a location accessed by said application, said location being within a computer network incorporating said server.
[0021] In accordance with another aspect of the present invention there is disclosed a server operating within a computer network, said server having at least one user browser application associated therewith and via which said user browser application accesses locations within said network, said server comprising:
[0022] means for receiving from said user browser application, bookmark information relating to a location recorded for subsequent access by said user browser application;
[0023] means for integrating said bookmark information received from plural ones of said user browser applications to form a database of said bookmark information.
[0024] In accordance with another aspect of the present invention there is disclosed a method of facilitating access to multimedia content distributed over a computer network, said method comprising the steps of:
[0025] at a server within said network, monitoring network location recording activities of subscribers of said server to form a ranked list of network identifiers recorded by said subscribers, said network identifiers being a source of multimedia content;
[0026] receiving search criteria relating to said multimedia content entered by a subscriber;
[0027] using said ranked list to determine a subset thereof;
[0028] applying said received search criteria to said determined subset; and
[0029] providing, to said subscriber of said server, results of applying said criteria to said determined subset, said results being provided according to an ordering of the corresponding originating network identifier in said list.
[0030] Other aspects of the present invention, including apparatus and systems for implementing the methods and computer program product related thereto are also disclosed
BRIEF DESCRIPTION OF THE DRAWINGS
[0031] One or more embodiments of the present invention will now be described with reference to the drawings and appendices in which:
[0032] FIG. 1 is a block diagram showing the operating environment of a multimedia access system;
[0033] FIG. 2 is a more detailed block diagram showing how the media browser of FIG. 1 accesses metadata databases;
[0034] FIG. 3 is a flow chart depicting the communication process between the media browser and the metadata server;
[0035] FIG. 4 shows the visual appearance of the user interface of the media browser component of the multimedia access system;
[0036] FIG. 5 is a flow chart showing the preferred browsing process of the media browser;
[0037] FIG. 6 is a flow chart showing the preferred searching process of the media browser;
[0038] FIG. 7 depicts a structured image metadata database;
[0039] FIG. 8 shows and example of XML metadata that is used by the for browsing process;
[0040] FIG. 9 is a schematic block diagram representation of a computer system upon which the media browser may operate;
[0041] FIG. 10 depicts an example implementation of the system of FIGS. 1
to 8;
[0042] FIG. 11 depicts customisation of the media browser for different devices;
[0043] FIG. 12 shows an arrangement by which the right to use multimedia content may be controlled;
[0044] FIGS. 13A, 13B and 13C illustrate methods by which metadata links may be communicated between devices;
[0045] FIG. 14 shows an arrangement by which a current media browser session may be switch from one device to another;
[0046] FIG. 15 depicts an example of how a source description can be transformed into a normalised description that is presentable by the media browser arrangement,
[0047] FIG. 16 is an illustration of a preferred media browser graphical user interface;
[0048] FIGS. 17A and 17B illustrate how "breadcrumb" navigation is used in the interface of FIG. 16;
[0049] FIGS. 18A to 18D depict system user via a further modified media browser graphical user interface;
[0050] FIG. 19 is a depiction of another modified graphical user interface;
[0051] FIG. 20 is a flowchart of using bookmarking for focussing searching activities;
[0052] Appendix 1 is an XML source description for the example of FIG. 15; and
[0053] Appendix 2 is an XML stylesheet which may be used to transform the source description of Appendix 1 to a normalised form.
DETAILED DESCRIPTION
[0054] I. Overview
[0055] FIG. 1 shows a multimedia access system 100 in which a computer application program, hereinafter called a media browser 101, operates on a local computer 105 to form a connection to a computer network, such as the Internet 102. As illustrated, the Internet 102 has associated therewith a number of server computers 108 and 109, each of which may host a number of Web sites and for each of which there is a corresponding store 112 and 114 in which multimedia content may be retained. The local computer 105 similarly may also have an associated store 107, although such is not essential to the implementation. The media browser application 101 provides a single user interface for a user of the local computer 105 to browse and search the system 100 for multimedia items using electronically-accessible metadata. In other words, the media browser 101 operates on metadata. Any playing/viewing of multimedia content is achieved by the use of plug-in media tools and is separated from the metadata-related processing. The media browser 101 is described in more detail in Section IV below.
[0056] The described arrangements may practiced using a general-purpose computer system 900, such as that shown in FIG. 9 wherein the processes of FIG. 1 and to be described are implemented as software, such as an application program executing within the computer system 900. In particular, the method of media browsing is effected by instructions in the software that are carried out by the computer system The software for the media browser 101 may be divided into essentially two separate parts; one part for managing the browsing and searching requests for a user, and another part to manage the user interface between the latter and the user. These two parts can be executed on separate computers in a mode known as client-server mode. The software may be stored in one or more computer readable media, including the storage devices described below, for example. The software is loaded into computers of the system from the computer readable media, and then executed by the computers. A computer readable medium having such software or computer program recorded thereon is a computer program product. The use of the computer program product in a computer preferably effects an advantageous apparatus for media browsing.
[0057] The computer system 900 comprises a computer module 901, input devices such as a keyboard 902 and mouse 903, output devices including a printer 915 and a audio-visual output device 914. A Modulator-Demodulator (Modem) transceiver device 916 is used by the computer module 901 for communicating to and from a communications network 920, for example connectable via a telephone line 921 or other functional medium. The network 920 may for example be the Internet, and/or other network systems, such as a Local Area Network (LAN) or a Wide Area Network (WAN). Collectively, the devices 901-916 may form for example either one or any of the local computer 105 or server computers 108 and 109 of FIG. 1 and are often are described as computer workstations
[0058] The computer module 901 typically includes at least one processor unit 905, a memory unit 906, for example formed from semiconductor random access memory (RAM) and read only memory (ROM), input/output (I/O) interfaces including a audio-visual interface 907, and an I/O interface 913 for the keyboard 902 and mouse 903 and optionally a joystick (not illustrated), and an interface 908 for the modem 916. A storage device 909 is provided and typically includes a hard disk drive 910 and a floppy disk drive 911. A magnetic tape drive (not illustrated) may also be used. A CD-ROM drive 912 is typically provided as a non-volatile source of data. The components 905 to 913 of the computer module 901, typically communicate via an interconnected bus 904 and in a manner which results in a conventional mode of operation of the computer system 900 known to those in the relevant art. Examples of computers on which the described arrangements can be practised include IBM-PC's and compatibles, Sun Sparcstations or alike computer systems evolved therefrom.
[0059] Typically, the application program is resident on the hard disk drive 910 and read and controlled in its execution by the processor 905. Intermediate storage of the program and any data fetched from the network 920 may be accomplished using the semiconductor memory 906, possibly in concert with the hard disk drive 910. The audio-visual output device 914
may be used to provide a graphical user interface to the application program by which user input may be afforded via the keyboard 902 and by clicking buttons on the mouse 903 as a mouse-cursor is manoeuvred across the interface represented on the audio-visual output device 914. In some instances, the application program may be supplied to the user encoded on a CD-ROM or floppy disk and read via the corresponding drive 912 or 911, or alternatively may be read by the user from the network 920 via the modem device 916. Still further, the software can also be loaded into the computer system 900 from other computer readable medium including magnetic tape, a ROM or integrated circuit, a magneto-optical disk, a radio or infra-red transmission channel between the computer module 901
and another device, a computer readable card such as a PCMCIA card, and the Internet and Intranets including e-mail transmissions and information recorded on websites and the like. The foregoing is merely exemplary of relevant computer readable media. Other computer readable media may be practiced without departing from the scope and spirit of the invention.
[0060] Returning to FIG. 1, the metadata used by the media browser 101 can be accessed directly from the local computer 105 or from any accessible site on the Internet 102 such as the server 108. Typically the metadata for a collection of multimedia content is stored in collections (eg. repositories or databases) with each item of content having at least one corresponding metadata item. As seen in FIG. 1, each content database or store 107, 112 and 114 has associated therewith a corresponding database 106, 110 and 111 respectively, which is configured to retain metadata items to facilitate access to the content within the corresponding respective content database or store 107, 112 and 114. Hereinafter, a metadata item is also referred to as a description of its corresponding item (typically, of content) and the term metadata collection refers to collections of such descriptions.
[0061] In the preferred arrangement, the media browser 101 is able to access the metadata without having to access the content (107, 112, 114). In other words, a description is not stored as an integral part of an item of content. This means that the media browser 101 does not need to be able to directly interpret the large number of storage/transport formats for audiovisual content in order to access metadata.
[0062] The media browser 101 assumes that each description (in the databases 106, 110 and 111) has a link to its corresponding content in the content database or store (107, 112, 114). If the content is stored electronically, these links can be actuated or electronically followed (eg. 120, 115, 116) by a user or process. Alternatively links, such as the link 118, can describe a route to a non-electronic location (eg. a film archive). Non-electronic links are not active (ie. unable to be followed by a remote user or process) and hence are only informative of available content. Accordingly, with such non-electronic links, remote users may not have a capacity to preview content using the media browser 101.
[0063] The media browser 101 requires that the metadata can be expressed in a standard manner. In the preferred arrangement, the syntax and structure of individual descriptions are determined by a schema. Descriptions of different items of content can use different schemas. Typically the schema used reflects the type of content and the typical use or purpose of the content. For example, a metadata schema for geological satellite images would most likely be significantly different to a schema for digital home video.
[0064] Schema may differ in their syntactical structure and the nature of the types of description components (hereinafter called descriptors). For example, a schema for digital home video may model descriptions of this type of content to contain a digital video tape, which contains one or more scenes, each of which contain one or more clips or shots. The geological satellite image schema may simply have a number of descriptors, with a particular geological focus, which are used to describe each image. In the preferred arrangement schemas are represented using the W3C Extensible Markup Language (XML) Schema language and individual descriptions are represented as XML documents. The metadata representation is described further in Section II.
[0065] FIG. 2 shows an example of how a media browser 101 can access metadata over the Internet 102. All access to metadata is achieved using links with the target of each link being expressed as a Uniform Resource Identifier (URI). These links can be actuated either automatically by the media browser 101 or in response to a user action (eg. clicking on an item of interest).
[0066] In the case where the metadata is stored in an XML repository (collection of XML documents) 200, the media browser 101 can provide access to the metadata stored in the repository 200 using a link to an XML description of the repository 200. This description represents the structure of the repository 200 that is presented to a user of the media browser 101. The XML description is represented in the same way as a description of a multimedia item of content. In other words, the description preferably conforms to an XML schema that is accessible to the media browser 101, and which describes the structure of the repository 200. The XML description can contain links to other descriptions of particular sections of the repository 200 (in other words, the description of the repository 200 does not need to be contained within a single XML document). Ultimately the repository XML description has links to descriptions of multimedia items. Each description of a multimedia item in the repository 200 preferably contains a link 201 to a corresponding multimedia item in the corresponding content collection 202. This enables the media browser 101
to be able to retrieve these items if a user or customer selects to view or play the item based on the presented metadata.
[0067] In the case where access to a non-XML repository, here called a legacy database 210, is desired, the link described above with reference to FIG. 1 must operate through a server module called a metadata server 212. The metadata server 212 is preferably located, though not necessarily, at the site of the metadata (ie. either local or remote) and is configured and controlled by the owner of the metadata. The purpose of the metadata server 212 is to effectively translate the metadata stored in a legacy database 210 to the format required by media browser 101. In other words, the metadata server 212 should preferably provide access to one or more schemas for the metadata and dynamically generate XML descriptions that conform to these schemas. Preferably, a metadata server 212 need only provide schema definitions that describe the structure/syntax of the metadata collection, and the structure/syntax of the individual descriptions stored in the legacy database 210. These schema definitions may be contained in one or more XML Schema documents. As with the case where the remote metadata is stored in an XML repository 200, the descriptions of multimedia items, that the metadata server 212
generates, contain links to the corresponding multimedia items stored in a content collection 214 corresponding to the legacy database 210.
[0068] A link to a metadata server is also represented using a URI. The URI is composed of a network identifier component, which is a URI itself, and a query string which specifies details of the metadata server request. The request can be executed using Hypertext Transfer Protocol (HTTP) over the Internet. Processing of the query results in descriptions of either the structure of the collection or multimedia items depending on how the metadata server 212 interprets the query string.
[0069] The descriptions that are dynamically generated by the metadata server 212 can be in response to media browser user browsing or search requests. Metadata servers are discussed further in Section III below.
[0070] II. Metadata Representation
[0071] The preferred arrangement assumes that all descriptions of multimedia items conform to a schema, and that schemas are expressed or represented using the W3C schema language, XML Schema. Individual descriptions are represented using XML document instances. XML Schemas are also represented as XML documents. Therefore descriptions (eg. of multimedia items) can be stored along with their respective schemas in XML repositories or object stores. Alternatively, the descriptions can be stored in a database and effectively translated into XML documents when required.
[0072] Each description contains reference(s) to schemas to which it conforms. Each reference is expressed using a URI (eg. http://somesite/schemas/DigitalVideoSchema.xsd). This means that once a media browser has access to a description it can directly access the schema or schemas to which the description conforms.
[0073] Technically, each XML element in a description (XML document) is declared to belong to a uniquely identified namespace. The XML document can then provide a hint to a processor, using the attribute schemaLocation (in the XMLSchema-instance namespace), for the location of a schema that contains definitions for a particular namespace. Hence an XML document, and thus also a description, indirectly rather than directly references one or more schemas.
[0074] In this document, the term "descriptor" is used to refer to a component, or atom, of a description. Each descriptor comprises a feature (descriptor name) and a value (description value). In some cases, the descriptor value comprises other descriptors, and thus may form a "complex descriptor". In other cases, the descriptor value is a scalar value such as a string or date (ie. simple or atomic descriptors). In all cases media browser 101 assumes that descriptors are represented with the element (tag) name being the descriptor name and the content of the element being descriptor value. For example, a simple descriptor may use the textual content of the element (ie. the text between the tags) to represent the value of the descriptor (eg. a date, text string, enumeration, etc.).
[0075] This assumption about the structure of the metadata is not unlike how many practitioners currently use markup languages. In other words, it does not require significant changes from how practitioners might represent particular metadata vocabularies.
[0076] Some examples of descriptors are now provided. In the simple descriptor, <Photographer>John Smith</Photographer>, Photographer is the name of the descriptor and John Smith is the value of the descriptor. The type of the text of a simple descriptor can be constrained using the simpleType construct of XML Schema.
[0077] In the example shown in FIG. 8, both VideoScene and Clip are complex descriptors. The value of the VideoScene descriptor is the markup that is contained within the start and end tags of the descriptor. The name of the descriptor is the tag name (ie. VideoScene). Similarly the value of the Clip complex descriptor is that markup contained between the start and end tags of the Clip descriptor. The Clip descriptor value contains two simple descriptors, Date and Location. The value of the Location descriptor is the text contained between the start and end Location tags (ie. Sydney, Australia).
[0078] In order to be able to better interpret the basic semantics of descriptions for the purposes of visually presenting descriptions in a meaningful way to users, the preferred arrangement includes a core schema which contains definitions of a number of basic attributes that description schema designers can use when they define their descriptors. An example of the definitions included in this core schema are shown below as Example A, in which only a fragment of the actual schema is shown. The definitions in this example belong to a unique media bowser namespace which is identified in the example by the prefix "mb:". It is assumed that this prefix assignment has been performed earlier in the schema using an xmlns attribute.
1EXAMPLE A 1. <simpleType name = `DescriptorType`> 2. <restriction base = `string`> 3. <enumeration value = `TOC`/> 4. <enumeration value = `Index`/> 6. <enumeration value = `Other`/> 6. </restriction> 7. </simpleType> 8. <attribute name = `Id` type = `ID` /> 9. <attribute name = `textIdentifier` type = `string` /> 10. <attribute name = 1 visualIdentifier` type = `anyURI`/> 11. <attributeGroup name = `DescriptorAttributes`> 12. <attribute ref = `mb:id` /> 13. <attribute ref = mb:textIdentifier` /> 14. <attribute ref = mb:visualIdentifier` /> 15. <attribute name = `updateable` type = `boolean` default = `false` /> 16. </attributeGroup> 17. <attributeGroup name = `TOCDescriptorAttributes`> 18. <attributeGroup ref = `mb:DescriptorAttributes` > 19. <attribute name = `descriptorType` type = `mb:DescriptorType` fixed = `TOC` /> 20. </attributeGroup> 21. <attributeGroup name = `IndexDescriptorAttributes`> 22. <attributeGroup ref = `mb:DescriptorAttributes` /> 23. <attribute name = `descriptorType` type = `mb:DescriptorType` fixed = `Index`/> 24. </attributeGroup>
[0079] The attribute descriptorType is used to define whether a descriptor is to be treated as part of a Table of Contents (TOC descriptor), or as part of an index (index descriptor).
[0080] A TOC descriptor is used to describe the structure of a description, and is typically a complex descriptor. A TOC descriptor is navigable in the sense that such must contain a link within either its attributes or within the attributes of its children. The target of the link can be either a further description or an item of content. A TOC descriptor is similar to an entry in a table of contents of a book in that it enables a reader to go directly to a section of the work.
[0081] Index descriptors are typically leaf nodes of a hierarchically-composed descriptor structure and are often referred to as properties (ie. the type of descriptive information that is displayed using a properties dialog in a Microsoft Windows (registered trade mark) system. Section IV below describes how the descriptorType attribute is used by the media browser.
[0082] Attributes are also used to contain visual and/or textual identifiers for a descriptor. A visual identifier (ie. visualIdentifier attribute) can be the URI of a thumbnail or movie/audio track preview. A text identifier (ie. textIdentifier attribute) can be used in the place of, or in addition to, a visual identifier. A text identifier typically contains a string value which describes the descriptor. In the absence of a visual identifier, the media browser can construct a visual representation based on this text value. These core attributes "drive" the user interface of the media browser. In other words, they have been included for presentation purposes.
[0083] In addition to these visualisation attributes that are defined in the core schema, the preferred arrangement uses the linking attributes of the W3C XLink Recommendation (as described at http://www.w3.org/TR/xlink ) to provide linking semantics. XLink provides a framework for creating both basic unidirectional links, such as the HTML <A> linking element, and more complex linking structures. Simple linking elements are a common linking requirement for the preferred arrangement. These links can be used to represent links between two descriptors (ie. items of metadata) and links between descriptors (metadata) and content (eg. images, video, etc.). XLink also provides for other linking types such as extended links, locators and arc. The full list of linking types is described at http://www.w3.org/TR/xlink.
[0084] The existence of a link, using XLink, is asserted by an XLink linking element. These elements need to be recognised by applications in order to provide appropriate display or behaviour. XLink uses a namespace to accomplish link recognition. The XLink namespace used by the preferred arrangement has the URI, http://www.w3.org/1999/xlink, and is associated with the xlink prefix. This association is achieved using the xmlns attribute of XML (eg. xmlns:xlink=`http://www.w3.org/1999/xlink`). XLink's namespace provides definitions of global attributes that can be used on elements that are in any arbitrary namespace. These global attributes (xlink:type, xlink:href, xlink:role, xlink:title, xlink:show, xlink:actuate, xlink:from and xlink:to) can be used to make elements recognisable as linking elements. For example, if the value of the xlink:type attribute is set to "simple" for a particular element, then that element is treated as a simple linking element and the value of the attribute, xlink:href, contains the target of that link. For the purposes of this description, definitions of the linking attributes using XML schema are included below in Example B.
2EXAMPLE B 1. <?xml version=`1.0`?> 2. <schema 3. xmlns = `http://www.w3.org/2001/XMLSchems` 4. xmlns:xlink = `http://www.w3.org/1999/xlink` 5. targetNamespace = `http://www.w3.org/1999/xlink` 6. attributeFormDefault = `qualified` 7. version = `1.0`> 8. <simpleType name = `LinkType`> 9. <restriction base = `string`> 10. <enumeration value = `simple`/> 11. <enumeration value = `extended`/> 12. <enumeration value = `locator`/> 13. <enumeration value = `arc`/> 14. <enumeration value = `resource`/> 15. <enumeration value = `title`/> 16. <enumeration value = `none`/> 17. </restriction> 18. </simpleType> 19. <simpleType name = `ShowType`> 20. <restriction base = `string`> 21. <enumeration value = `new`/> 22. <enumeration value = `replace`/> 23. <enumeration value = `embed`/> 24. <enumeration value = `other`/> 25. <enumeration value = `none`/> 26. </restriction> 27. </simpleType> 28. <simpletype name = `ActuateType`> 29. <restriction base = `string`> 30. <enumeration value = `OnLoad`/> 31. <enumeration value = `onRequest`/> 32. <enumeration value = `other`/> 33. <enumeration value = `none`/> 34. </restriction> 35. </simpleType> 36. <attribute name = `type` type = `xlink:LinkType` default `simple`/> 37. <attribute name = `show` type = `xlink:ShowType` default = `new`/> 38. <attribute name = `role` type = `QName` default = `resource`/> 39. <attribute name = `actuate` type = `xlink:ActuatType`/> 40. <attribute name = `href` type = `anyURI`/> 41. <attribute name = `arcrole` type = `string`/> 42. <attribute name = `title` type = `string`/> 43. <attribute name = `label` type = `NMTOKEN`/> 44. <attribute name = `from` type = `NMTOKEN`/> 45. <attribute name = `to` type = `NMTOKEN`/> 46. </schema>
[0085] A particular schema can use the core XLink and media browser attributes when declaring individual descriptors for a schema. In Example C below, the particular descriptors VideoClip, Date and Photographer are declared in a particular schema. Note that only a fragment of an actual schema is shown and reference to the media browser and XLink namespaces is assumed via the namespace prefixes mb and xlink, respectively. In XML Schema these namespace prefixes are assigned using the xmlns attribute of the XML Schema language. The media browser attributes are referenced unchanged from their definitions as seen at line 21 of Example C for TOCDescriptorAttributes. However one of the XLink attributes that is referenced, for example as seen at line 24, is further refined from its original definition. For example, the VideoClip descriptor is a simple linking element so the xlink:type attribute's value will assume the default value of "simple". With a supple link, the element (descriptor) is the link source and a single linkend must exist. This single linkend is represented using the xlink:href attribute. A value must be supplied for this attribute for the simple link to be valid (hence the use constraint for this attribute is set to "required"). Note also that the xlink:href attribute for the VideoClip descriptor will assume its default value of "resource" (ie. the target of the link should be assumed to be the item of content being described).
3EXAMPLE C 1. <element name = `VideoClip`> 2. <complexType> 3. <element name = `Date`> 4. <complexType> 5. <simpleContent> 6. <extension base = `date`> 7. <attributeGroup ref = `mb:IndexDescriptorAttributes`/> 8. </extension> 9. </simpleContent> 10. </complexType> 11. <element> 12. <element name = `Photographer`> 13. </complexType> 14. <simpleContent> 15. <extension base = `string `> 16. <attributeGroup ref = `mb:IndexDescriptorAttributes`/> 17. </extension> 18. </simpleContent> 19. </complexType> 20. </element> 21. <attributeGroup ref = `mb:TOCDescriptorAttributes`/> 22. <attribute ref = xlink:type`/> 23. <attribute ref = `xlink:role`/> 24. <attribute ref = `xlink:href` use = `required`/> 25. </complexType> 26. </element> A description conforming to this particular schema fragment may contain the fragment of Example D: Example D: 1. <VideoClip xlink:href = `http://someSite/content/video/clip999.mpg`> 2, <Date>2000-04-18</Date> 3. <Photographer>John Smith</Photographer> 4. </VideoClip>
[0086] In the preferred arrangement, the core media browser attributes are explicitly expressed in schemas and/or descriptions. Alterative arrangements can infer these attribute values from other information in descriptions, as described below. For example, a descriptor/element may be treated as part of the TOC if it contains a link within either its attributes or within the attributes of its children. Further, descriptors which do not have descendant links may be treated as index descriptors. Similarly visual identifiers may be automatically constructed from element (descriptor) names. Methods of inferring these attribute values are described in more detail later in this section.
[0087] Clearly there are alternative ways in which the core media browser and xlink semantics can be represented. For example, using XML Schema it is possible to define a core Descriptor type and from that core type, derive TOCDescriptor and IndexDescriptor types (see Example E below). Then individual schema definitions may extend these base types to provide implementation-based descriptors such as those defined in Example C. Also, although the preferred arrangement uses XML Schema as its schema representation languages, other suitably expressive schema languages may be used.
4EXAMPLE E 1. <simpleType name = DescriptorType`> 2. <restriction base = `string`> 3. <enumeration value = `TOC`/> 4. <enumeration value = `Index`/> 5. <enumeration value = `Other`/> 6. </restriction> 7. </simpleType> 8. <complexType name = `Descriptor`> 9. <attribute name = `Id` type = `ID`/> 10. <attribute name = `textIdentifier` type = `string`/> 11. <attribute name = `visualIdentifier` type = `anyURI`/> 12. <attribute name = `descriptorType` type = `DescriptorType`/> 13. <attribute name = `value` type =``xsd:anyType`> 14. <attribute ref = `xlink:href`/> 15. </complexType> 16. <complexType name = `TOCDescriptor`> 17. <restriction base = `Descriptor` 18. <attribute name = `id` type = `ID`/> 19. <attribute name = `textIdentifier` type = `string`/> 20. <attribute name = `visualIdentifier` type ==`anyURI`/> 21. <attribute name = `descriptorType` type = `DescriptorType` fixed = `TOC`/> 22. <attribute name = `value` use = `prohibited`/> 23. <attribute ref = `xlink:href`/> 24. </restriction> 25. </complexType> 26. <complexType name = `IndexDescriptor`> 27. <restriction base = `Descriptors`> 28. <attribute name = `id` type = `ID`/> 29. <attribute name = `textIdentitier` type = `string`/> 30. <attribute name = `visualIdentifier` type = `anyURI`/> 31. <attribute name = `descriptorType` type = `DescriptorType` fixed = `Index`/> 32. <attribute name = `value` type = `xsd:anyType`/> 33. <attribute ref = `xlink:href` use = `prohibited`/> 34. </restriction> 35. </complexType>
[0088] Interpretation of Metadata
[0089] In practice, not all the metadata that a user wishes to visualise using the media browser 101 will explicitly use the media browser 101 and the XLink attributes described above. On parsing a new description, the media browser 101 first attempts to identify the type of metadata that has been received, examples of which may include Dublin Core, MPEG-7 or DIG35 (for images). Each of these metadata standards are known in the art. Typically, this can be achieved by examining either the root element of the description or the namespace declarations. If the media browser 101 identifies a metadata standard, then the media browser 101 uses an XSLT stylesheet to transform the incoming document tree (description) into one that explicitly uses the media browser and Xlink attributes. No further processing is required. In other words, it is assumed that the transform results in a description that the media browser can present without further processing.
[0090] For all other descriptions, a check is performed to attempt to ensure that the preferred media browser attributes are present. If they are absent then the preferred implementation uses a list of rules for the creation of appropriate media browser attributes for the incoming metadata The rules are as follows:
[0091] (i) An href attribute is assumed to represent the target of a simple link and is represented as an xlink:href attribute. If the target value of the link is a URI with an extension of XML or no extension, then a link to another description is assumed (ie. xlink:role is set to `description`), otherwise the link is assumed to be a link to the relevant content (ie. xlink:role is set to `resource`). The type of the link is assumed to be simple (ie. xlink:type is set to `simple`).
[0092] (ii) An element is classified as a TOC descriptor if either the descriptor or any of its children contains a link (ie. mb:descriptorType is set to `TOC`). The link may be represented in the original metadata as element content or an attribute. An element not classified as a TOC descriptor is assumed to be an Index descriptor.
[0093] (iii) If a descriptor does not have a visualIdentifier or a textIdentifier then a textIdentifer is created with a value that is obtained either from a name attribute of the descriptor, if it exists, or from the element name. In this regard, the media browser 101 preferably always displays a visualIdentifier if one exists, otherwise the textIdentifer is used.
[0094] (iv) If a descriptor does not have a visualIdentifier but it does contain an attribute or child element having a name that indicates that it may be able to act as a visualIentifier (eg. keyFrame, thumbnail, preview, etc) then a visualIdentifier attribute is created using that attribute or element's value. This rule can be implemented by checking each attribute name against a list of possible visualIdentifier names.
[0095] Whilst the above lists only four rules, it will be appreciated that alternate and/or additional rules may be developed to provided for meaningful interpretation of unknown metadata types.
[0096] However, the use of an XSLT stylesheet is the desired approach when possible because a priori knowledge of the metadata format enables a stylesheet author to define informed transforms. For example, the value of the visualIdentifier attribute may be taken directly from the value of another attribute. An example of a transform for some arbitrary video metadata that is based on a subset of known extended Dublin Core attributes to a form useable by Media Browser is shown in FIG. 15.
[0097] In FIG. 15 the source and transformed descriptions are depicted as XML element node trees with attributes shown in the boxes to the right of the corresponding node. Elements are represented using the oval shapes. So, for example, in the source description 1580, the VideoDocument element 1500 has five attributes 1502, namely DC.Title, DC.Creator, DC.Subject, DC.Type, and href. The notation {att_name} is used to denote the value of the attribute of the corresponding element in the source document with the name att_name. The avptr notation is a method of addressing into audiovisual content using XPointer fragments. For example:
[0098] http://../AusWild883.mpg#avptr(time::2:05.00,2:55.20)
[0099] refers to the fragment of the audio visual content AusWild883.mpg, where the fragment starts 2 minutes and 5 seconds from the start of the content, and ends at 2 minutes and 55.2 seconds.
[0100] An XSLT transform 1528 of FIG. 15 is configured with knowledge of the syntax and semantics of a source description 1580 for a video document description 1500 having a number of attributes 1502 (eg. DC.Title). For example, the shown transform assumes that the value of the DC.Identifier attribute in the set of attributes 1510 of the source Scene elements 1504, 1506, 1508, and the DC.Identifier attribute in the set of attributes 1518 of the Shot elements 1512, 1514, 1516 is just a reference identifier and does not provide additional information. For this reason, the transform uses these references as the values of the mb:id attribute. If these identifiers did carry significance to the user of the metadata then these attributes could have been transformed into index descriptors as, for example, the DC.Description attribute of the Scene element 1544. Note also that in FIG. 15 the transformed description does not maintain the initial frame granularity of the source description. In other words, the normalised description 1530 does not contain Frame description as in the source description 1500. This represents a decision made by the designer of the stylesheet 1528 which typically operates with knowledge of the media browser interface 101.
[0101] In the example of FIG. 15, it may initially appear counterproductive to transform a description that uses elements to represent structure and attributes to represent properties, into an element tree. However, concepts of what information should be represented as attributes and what information should be represented as elements often vary with media type, as described above. For this reason, transforming source metadata into an element tree is a form of normalising the metadata, and the transform 1528 thus results in a normalised description 1590 able to be processed and presented by the media browser 101.
[0102] The source description 1580 is an XML document seen in Appendix 1, The media browser 101 does not attempt to transform any relevant schema, if one exists. Consequently, the transformed description does not conform to a schema and therefore the description cannot be annotated. This is emphasised in the transformed description by setting the updateable attribute of the media browser 101 to false in the root element 1532 of the transformed description 1590. The XSLT stylesheet used to achieve the transform 1528 is seen in Appendix 2.
[0103] III. Metadata Servers
[0104] A link to a metadata server 212 is represented using a URI. An expression describing the request is appended to a URI that uniquely identifies the metadata server 212. For example, the URI: http://somesite/myMetadata/Svr?<query_string>, has an identifier component which is the part of the URI preceding the question mark symbol and a request component which carries information about the request to be sent to the metadata server 212. The identifier component is itself a URI.
[0105] The preferred arrangement interprets the link by first using the identifier part of the URI to locate the metadata server 212 on the network 102. Failure to identify the metadata server 212 results in a failed link and the media browser 101 user can be notified of the failure to detect a Hag process. In the preferred arrangement, the metadata server 212 must be running as a process and the process being run by the metadata server 212 cannot be initiated from the media browser 101. In alternative arrangements, the media browser 101 may be configured to initiate the one or more metadata server processes.
[0106] When an identified metadata server 212 receives a request, the server 212 interprets the request and replies with an XML description that satisfies the request. Preferably the description is sent as XML, however the description may be encoded if desired or necessary. The types and elements used in the description are preferably defined in a schema that the media browser 101 can access. Although, the descriptions are not validated against their schema by the media browser 101 in the described arrangement, the media browser 101 prefers to have access to the schema. If a schema is not available then some media browser functions may not be available. Preferably, the types and elements of the schema used by the metadata server 212 are derived using the core attributes defined above in Section II.
[0107] The requests directed at the metadata server 212 may be for metadata required for browsing or a search expression. The request can also specify various parameters that control the delivery of the XML back to the requesting media browser service.
[0108] The results of requests that are directed at a metadata server 212
are descriptions which are preferably contained in an element, either of the type, or derived from the type MetadataCollection, an example of which is provided below as Example F The MetadataCollection type provides a means for the metadata server to explicitly return information to the requesting media browser application or service (eg. the number of items that satisfy the request and the number of items that are actually returned in the description).
5EXAMPLE F 1. <complexType = `MetadataCollection`> 2. <attribute name = `descriptorType` type = mb:DescriptorType fixed = `Other`> 3. <attribute name = `requestID` type = `string`/> 4. <attribute name = `noItemsIdentified` type = `Integer`/> 5. <attribute name = `noItemsReturned` type = `integer"> 6. <attribute name = `startItemReturned` type = `integer`/> 7. </complexType>
[0109] Before details of the request syntax are described, the overall processing model for the communication performed by the media browser 101
with a metadata server 212 is described with reference to the flow chart of FIG. 3. Firstly, in step 300, the metadata server 212 is identified from the URI. The request is then sent to the identified metadata server 212 in step 301. Specifically, what occurs in the preferred arrangement is that the URI containing the metadata server request is fetched using the HTTP. In other words, steps 300 and 301 are performed as a single process. The system then waits in step 302 for a reply. A check is performed in step 303 to see if a reply has been received. If not, then the waiting period is compared with a predetermined timeout in step 304
and if the waiting period is not greater than the timeout, control passes back to step 302. If the waiting period is greater than the timeout, an error is reported to the media browser user in step 306 and the process terminates in step 310 (ie. the metadata server 212 has not been reached for some reason).
[0110] If a reply is received in step 303 the media browser 101 examines the response. If the media browser 101 cannot process the response (eg. the response is not correctly structured) then an error is reported in step 306 and the process terminates in step 310. If the response is able to be processed (ie. parsed) then it is passed to the appropriate module in the media browser 101 for further processing and the process terminates in step 310.
[0111] The request syntax will now be discussed in more detail.
[0112] Typically roost legacy databases store metadata in relational databases and access these databases using Standard Query Language (SQL). On the other hand, XML documents, and hence the media browser 101, represent information (metadata) in an hierarchical fashion. The metadata server 212 request must provide a bridge between the two different representations. Although it may be simpler to implement metadata servers if the request was based on SQL, the media browser 101 uses XML-related technology. In particular, the metadata server request is based on the W3C Recommendation XPath Version 1.0, which may be found at http://www.w3.org/TR/xpath. It may also be possible to use the emerging W3C standard XQuery.
[0113] XPath provides an extremely understandable way to describe a class of nodes which are to be processed. It is declarative rather than procedural and uses a simple pattern syntax modelled after directory notation. The most common form of XPath expressions are location paths. A location path selects a set of nodes relative to a context node. A location path can be absolute (starts with a `/` to denote the root node) or relative (to a context node). For example, the expression book/author is a relative location path which selects all author children of book children of the context node. The XPath syntax is most easily understood by way of examples and examples are provided at http://www.w3.org/TR/xpat- h. A number of XPath examples are as follows:
6
(i) /* selects all the children of the root node (ii) /doc/chapter[5]/section[2] selects the second section of the fifth chapter of the doc. (iii) */para selects all the para grandchildren of the context node (iv) para[@type="warning"] selects all the para children of the context node that have a type attribute with the value of warning. (v) chapter[title="Introduct- ion"] selects the chapter children of the context node that have one or more title children with string value equal to Introduction.
[0114] The location path syntax of XPath is directly useable for representing browsing requests and also for structured queries. In order to package unstructured queries (search expressions) as requests to the metadata server, XPath's function notation is used. This requires a more detailed understanding of XPath.
[0115] The primary syntactic construct in XPath is the expression. An expression is evaluated to yield an object which is one of the following four basic types:
[0116] Node-set (an unordered collection of nodes without duplicates);
[0117] Boolean (true or false);
[0118] Number (a floating point value); and
[0119] String.
[0120] A location path, as discussed above, is a special case of an XPath expression. A location path returns the set of nodes selected by the path. The part of the location path that is enclosed by square brackets `[ ]` is called the predicate. The predicate is itself an XPath expression which returns a Boolean result which serves to filter the node set selected with respect to the defined axis (tree relationship between the nodes selected and the context node) of the location step.
[0121] An expression can also be a function call, which, optionally, takes arguments. The EBNF (Extended Backus Naur form) definition of a function call is taken from Section 3.2 of the above referenced W3C Recommendation found at http://www.w3.org/TR/xpath. An example of such an expression is:
[0122] FunctionCall ::=FunctionName `(`(Argument (`,`Argument)*)?`)`
[0123] Argument ::=Expr
[0124] Note the production Expr is the basic construct of XPath. A core function library exists which must be implemented by XPath implementations. Each function in the library is specified using a function prototype which gives the return type, the name of the function and the type of arguments. Although no core functions exist that can be used to pass the request to perform an unstructured query, it is trivial to extend XPath by defining a user function.
[0125] Therefore, the syntax for requests is based on XPath with additional functionality to specify parameters that control the transmission of metadata to media browser. The syntax is detailed below using EBNF:
[0126] Request ::=XPathExpression (`&`ParameterList)?
[0127] ParameterList ::=MaximumItems? (`&` StartItem)? (`&` NumberLevels)? (`&`TransactionID)?
[0128] MaximumItems ::=`maxItems=` Number
[0129] StartItem ::=`startItem=`Number
[0130] NumberLevels ::=`noLevels=`Number
[0131] TransactionID ::=`requestID=`Nmtoken
[0132] Number ::=Digit (Digit)*
[0133] The Request contains a single XPathExpression followed by an optional ParameterList. The XPathExpression matches the production LocationPath of the XPath Version 1.0 described at http://www.w3.org/TR/xpath with the exception that the predicate expression must support the following additional function call:
Function: Boolean query(unstructuredQuery)
[0134] This function can be included in a location path and can be used to request that the metadata server 212 pass the unstructured query on to a search engine associated with the database 210. For example, the location path /Lifestyles/images[query("surfing")] would therefore be interpreted by a metadata server 212 as finding all those images that are children of the Lifestyles node that satisfy the unstructured query "surfing". The query function makes imprecise queries possible. Note that the expression unstructuredQuery must be encoded appropriately for inclusion in the URI. Appropriate encoding is specified by the Network Working Group's Request for Comments (RFC) 2396 available from http://www.ietf.org/rfc.html.
[0135] Both Nmtoken and Digit mentioned above are defined in the XML Version 1.0 Recommendation (see http://www.w3.org/TR/1998/REC-xml-1998021- 0).
[0136] The ParameterList component of a Request is optional. ParameterList contains the optional individual productions MaximumItems, StartItem, NumberLevels and TransactionID which specify the maxItems, startItem, noLevels and requestID parameters, respectively. If any of these parameters are not specified then the media browser 101 uses a default value.
[0137] The parameter maxItems refers to the maximum number of items to be returned by the metadata server 212. So, for example, if a particular section of a collection contained a large number of items then media browser could request the first, say (n=101) items. The default value is specified by the user within the media browser 101. This parameter is automatically inserted into the Request by the media browser 101. If the user does not specify a value, a system default is used (eg. maxItems=100)
[0138] The startItem parameter allows the media browser 101 to get the next n items starting from a specified item number. The startItem parameter is useful in retrieving search results from a metadata server 212. If it is not specified in a URI, then a value of `1` is assumed by the metadata server 212.
[0139] The parameter noLevels enables the media browser 101 to define the structure of the returned description, Typically a single (hierarchical) level of description is required, however more levels may be desirable in the event of a user requesting a particular view that contains more than one level of the hierarchy (eg. scenes and clips for a video). If this parameter is not specified then a value of one (hierarchical) level is assumed.
[0140] The requestID parameter allows a request to be formulated that refers to a previous request. For example, it may be desirable to obtain the next set of items from a previous request. If a requestID is specified then the metadata server 212 will attempt to reply using the previous request that is identified by the requestID. If the request identified by the requestID is no longer available in a cache of the metadata server 212 then the processing associated with the request will have to be repeated. The requestID is a unique value for the metadata server 212 and is generated by the metadata server 212 (and may be based on a timestamp representing the receipt of the request by the metadata server 212). The requestID can be returned to the media browser 101 using an element having the type, or being derived from the type, MetadataCollection (see Example F).
[0141] Browsing Requests
[0142] In one implementation a default Request, used when initially gaining browsing entry to a metadata collection for the purposes of browsing, may be the XPathExpression, "/*" with any desired parameters formatted in a ParameterList (eg. "/*&maxItems=100&noLevels=2"). The corresponding URI would then be:
[0143] http://mySite/myMetadataSvr?/*&maxItems=100&noLevels=2
[0144] where //mySite/myMetadataSvr is the URI of the metadata server process.
[0145] On receipt of this request, the metadata server 212 invokes a procedure to satisfy the request. This procedure results in the dynamic generation of an XML description of the associated metadata collection. This description thus reflects a structure by which the associated metadata collection may be browsed. It is common for the metadata collection to be stored in a database of some form. For example, the metadata server 212 may be configured to provide category or publisher sections for the collection so that users can more readily browse the metadata. Typically these categories are reflected in the schema used to describe the database items. Alternatively, the metadata server 212 may reply to the request from the media browser 101 by simply sending a list of all the separate items in the database.
[0146] For the purpose of describing a typical scenario of use, consider an image metadata database with the following structure. The database is composed of a number of categories including Lifestyles, Sports and Animals as illustrated in FIG. 7. Whereas the Lifestyles category has no further structure (ie. it is composed of solely images), the Sports category is structured further into subcategories and the Animals category is structured further into subcategories and then image classes. It is not important how this data is actually stored for the purposes of the present description.
[0147] There is no fixed way that a metadata server 212 may implement its translation facility for its metadata collection. One possible way is described below.
[0148] The metadata server 212 generates descriptions based on an XML schema definitions of types for Categories, Subcategories, Classes and Images. Typically these schema definitions reside in a single XML Schema document. Preferably these definitions use the core attributes of the media browser 101 and the global XLink attributes (see Section II above). A basic example of such definitions is provided below in Example G of an XML Schema. Note, that the definitions can use the xlink:show attribute to direct the media browser 101 to "embed" the target of the link at the source (ie. the description fragment generated by the metadata server 212
would be simply included as the content of the link source element). Definitions may also set this attribute value to "replace", in which case the media browser 101 would replace the descriptor, which is the link source, with the description fragment served by the metadata server 212.
7EXAMPLE G XML Schema Example 1. <?xml version=`1.0`22
2. <schema 3. xmlns = `http://www.w3.org/1999/XMLSchema` 4. xmlns:mb = `http://www.cisra.com.au/MediaBrowser` 5. xmlns:xlink = `http://www.w3.org/1999/xlink` 6. xmlns:image = `http://www.somesite/ImageLibrary` 7. targetNamespace = `http://www.somesite/ImageLibrary` 8. version = `1.0`> 9. <element name = `ImageLibrary`> 10. <complexType> 11. <complexContent> 12. <extension base = `mb:MetadataCollection`> 13. <choice> 14. <element ref = `im:Category` minOccurs = `0` maxOccurs = `unbounded`/> 15. <element ref = `im:SubCategory` minOccurs = `0` maxOccurs = `unbounded`/> 16. <element ref = `im:Class` minOccurs = `0` maxOccurs = `unbounded`/> 17. <element ref = `im:Image` minOccurs = `0` maxOccurs = `unbounded`/> 18. </choice> 19. </extension> 20. </complexContent> 21. </complexType> 22. </element> 23. </element name = `Category`> 24. <complexType> 25. <choice> 26. <element ref = `im:SubCatgeory/> 27. <element ref = `im:Image`/> 28. </choice> 29. <atiributeGroup ref = `mb:TOCDescriptorAttributes`/> 30. <attribute ref = `xlink:type`/> 31. <attribute ref = `xlink:href`/> 32. <attribute ref = `xlink:role`/> 33. <attribute ref = `xlink:show`/> 34. </complexType> 35. </element> 36. <element name = `SubCategory`> 37. <complexType> 38. <choice> 39. <element ref = `im:Class`/> 40. <element ref = `im:Image`/> 41. </choice> 42. <attributeGroup ref = `mb:TOCDescriptorAttributes`/> 43. <attribute ref = `xlink:type`/> 44. <attribute ref = `xlink:href`/> 45. <attribute ref = `xlink:role`/> 46. <attribute ref = `xlink:show`/> 47. </complexType> 49. </element> 50. <element name = `Class`> 51. <complexType> 52. <element ref = `im:Image`/> 53. <attributeGroup ref = `mb:TOCDescriptorAttributes`> 54. <attribute ref = `xlink:type`/> 55. <attribute ref = `xlink:href`/> 56. <attribute ref = `xlink:role`/> 57. <attribute ref = `xlink:show`/> 58. </complexType> 59. </element> 60. element name = `Image`> 61. <complexType> 62. <sequence> 63. <element ref = `im:ImageID`/> 64. <element ref = `im:Name`/> 65. <element ref = `im:Caption`/> 66
<element ref = `im:Photographer`/> 67. <element ref = `im;Keywords`/> 68. </sequence> 69. <attributeGroup ref = `mb:TOCDescriptorAttributes`/> 70. <attribute ref = `xlink:type`/> 71. <attribute ref = `xlink:href`/> 72. <attribute ref = `xlink:role`/> 73. <attribute ref = `xlink:show`/> 74. </complexType> 75. </element> 76. <element name = `Name`> 77. <complexType> 78. <simpleContent> 79. <extension base = `string` 80. <attributeGroup ref = `mb;IndexDescriptorAttributes`/> 81. </extension> 82. <simpleContent> 83. </complexType> 84. </element> 85. <element name = `Photographer`> 86. <complexType> 87. <simpleContent> 88. <extension base = `string2 > 89. <attributeGroup ref = `mb:IndexDescriptorAttributes`/> 90. </extension> 91. </simpleContent> 92. </complexType> 93. </element> 94. <element name = `ImageID`> 95. <complexType> 96. simpleContent> 97. <extension base = `string`> 98. <attributeGroup ref = `mb:IndexDescriptorAttributes`/> 99. </extension> 100. </simplecontent> 101. </complexType> 102. </element> 103. <element name = `Caption`> 104. <complexType> 105. <simpleContent> 106. <extension base = `string`> 107. <attributeGroup ref = `mb:IndexDescriptorAttributes`/> 108. </extension> 109. </simpleContent> 110. </complexType> 111. </element> 112. <element name = `Keywords`> 113. <complexType> 114. <simpleContent> 115. <extension base = `string`> 116. <attributeGroup ref = `mb:IndexDescriptorAttributes`/> 117. </extension> 118. </simpleContent> 119. /complexType> 120. </element> 121. </schema>
[0149] The schema document in Example G contains a declaration for root element, ImageLibrary, that extends the MetadataCollection type defined for the media browser (mb) namespace (see Example F). It thus inherits all the attributes defined for the base type (ie. descriptorType, requestID, noItemsIdentified, noItemsReturned, startItemReturned). In addition it is defined to contain any of the following list of descriptors: Category, SubCategory, Class or Image. What is actually returned by the metadata server as the content of this root element will depend on the request received.
[0150] The schema document also contains declarations for the following TOC descriptors Category, SubCategory, Class and Image. Each of these descriptors is defined to contain the attribute group TOCDescrptorAttributes (from the mb namespace and defined in Example A) and a set of linking attributes (type, href, role and show from the xlink namespace).
[0151] In this example, the type, show and role attributes will default to "simple", "new" and "resource", respectively, unless overwritten in an instance (eg. the XML document generated by the metadata server in response to a request). So, for example, the default value of the xlink:show attribute will need to be overwritten if a link to another metadata server request is to be included. In this case, usually the desired value for this attribute is "embed" which instructs the receiver of the generated description to embed the element content description as a child element of the descriptor containing the link source to the metadata server. It is also possible to set the value of the xlink:show attribute to be "replace" which means that the element content of the generated description should replace the descriptor containing the original link to the metadata server. The default value for the xlink:show attribute can be used if a link to a resource is the objective. In this case you want the resource to be displayed in a new window (hence the use of the word "new" for the default value).
[0152] Also the generated description will need to overwrite the value of the xlink:role attribute if the objective of link is to link to a further description. In this case the value of this attribute should be set to "description".
[0153] Each of the declared descriptors in Example C inherit a visualIdentifier attribute (from either the TOCDescriptorAttributes or IndexDescriptorAttributes group). This attribute is used by the media browser 101 to provide a visual representation of the content of the item. For example, if the item is an image then the visualIdentifier attribute value will typically contain the URI of a thumbnail of the image. In the case of categories, subcategories and classes the visualIdentifier attribute value can contain the URI of an icon. If this attribute is not specified then, preferably, the media browser 101
generates the visual identifier for the item from a provided textIdentifier attribute value, or, in the event that this value is also not provided, from the name of the element (in this case Image, Class, Subcategory or Category).
[0154] On receipt of the "/*" Request, the metadata server 212 generates an XML description of the collection as in the XML fragment below of Example H. The description is contained in an element declared to be of type MetadataCollection (see Example G) and it contains returning links to the metadata server for further descriptions. Note that the metadata server needs only specify the XPathExpression in its returning links. It is the responsibility of the media browser to add the ParameterList to the URI before despatching the request.
8EXAMPLE H Returned XML Description Fragment 1. <ImageLibrary 2. requestID = `19999123` 3. noItemsIdentified = `3` 4. startItemReturned = `1` 5. noItemsReturned `3`> 6. <Category 7. textIdentifier = `Lifestyles` 8. xlink:href = "http://mySite/myMetadaSvr?Category[@- textIdentifier= `Lifestyles`]/Image" 9. xlink:role = `description` 10. xlink:show = `embed` 11. visualIdentifier = `http://mySite/Metadata/icons/Lifestyles.gif`/> 12. <Category 13. textIdentifier = `Sports` 14. xlink:href = "http://mySite/myMetadaSvr?Category[@textIdentifier= `Sports`]/Subcategory" 15. xlink:role = `description` 16. xlink:show = `embed` 17. visualIdentifier = `http://mySite/Metadata/icons/Sports.gif`/> 18. <Category 19. textIdentifier = `Animals` 20. xlink:href="http://mySite/m- yMetadataSvr?Category[@textIdentifier `Animals`]/Subcategory" 21. xlink:role = `description` 22. xlink:show = `embed2
23. visualIdentifier = `http://mySite/Metedata/icons/Animals.gif/> 24. </ImageLibrary>
[0155] In the Example H above description, XPathExpressions in the return links to the metadata server 212 are used to identify links to each of the images in the Lifestyles category and the subcategories in the Sports and Animals categories. These links would be activated when a user selects to expand one of the above items when they were visually presented in media browser 101. In the preceding and following examples, the XPathExpressions have been specified as relative location paths assuming that the context node is the root node of the collection, Alternatively, absolute paths can be used.
[0156] In Example H above, the URI targets of the return links to the metadata server 212 contains the `[` and `]` characters. In general, according to RFC 2396, it is unwise to leave these characters unencoded in a URI because they can be excluded by some gateways and transport agents. The characters have been left unencoded in this and the following Examples for ease of reading.
[0157] If, for example, the visual identifier for the `Sports` category was selected when the XML fragment shown in Example H was processed and presented to the user, then the corresponding returning link to the metadata server would be actuated. The metadata server 212 would respond to this link by generating and returning a description fragment as now indicated below in Example I.
9EXAMPLE I Returned XML Description Fragment 1. <ImageLibrary 2. requestID = `19999124` 3. noItemsIdentified = `1200` 4. startItemReturned = `1` 5. noItemsRetured = `100`> 6. <Subcategory 7. textIdentifier = `Basketball` 8. xlink:href = "http://mySite/myMetadataSvr?Category[@textIdentifier= `Sports`]/Subcategory[@textIdentifier=`Basketball`]/Image" 9. xlink:role = `description` 10. xlink:show = `embed`> 11. <Subcategory 12. textIdentifier `Football` 13. xlink:href = "http://mySite/myMetadataSvr?Category]@textIdentifier= `Sports`]/Subcategory]@textIdentifier=`Football`]/Image" 14. xlink:role = `description` 15. xlink:show = `embed`/> 16. <subcategory 17. textIdentifler = `Hockey 18. xlink:href = "http://mySite/myMetadataSvr?Category[@textIdentifier= `Sports`]/Subcategory[@textIdentifler=`Hockey`]/Image" 20. xlink:role = `description` 21. xlink:show = `embed`/> 22. </ImageLibrary>
[0158] It is preferred that the returned description be well-formed. Further, the returned description must be able to be parsed by the media browser 101. The action of the media browser 101 on receiving the contents of a link depend on the xlink attribute show as described previously. Typically, this attribute will be set to "embed" in which case the received description is embedded at the source of the link. If the received description used a container element (eg. of type MetadataCollection as defined in Example F) then this element is also embedded. Preferably embedded container elements are defined as having a descriptorType value of "Other" (see Example A). Alternatively as previously mentioned, the xlink:show attribute can be set to "replace" in which case the contents of the link will replace the element containing the link source. If the xlink:show attribute is not included for the linking element in the description generated by the metadata server then the default action is "new". This means that the contents of the link are displayed in a new window. Clearly this is behaviour that is desirable for content (ie. a resource) rather than for a description.
[0159] The description of the collection may be further explored by a user selecting one of these subcategories. This action would result in the metadata server 212 generating a description of the images contained in the selected subcategory.
[0160] Note that the description of Example I, which is dynamically generated by the metadata server 212, contains only a single hierarchical level. This can be altered by specifying the noLevels parameter in the ParameterList of the URI. In some cases a Request might require two levels of hierarchical description in order to generate a view that requires both the parent and the children TOC elements. For example, if the media browser 101 was using a two-level view and wished to retrieve descriptions that contained two levels of TOC hierarchy, then the media browser 101 would append the "noLevels=2" parameter to the URI. For example, the link:
[0161] http:/gmySite/myMetadataSvr?Category/Subcategory(Categor /@textldentifier=`Sports`]&noLevels=2
[0162] would result in the description fragment shown below in Example J.
[0163] The second level is assumed to be the TOC children of the level targeted by the link. Preferably, when the value of noLevels is greater than one, the values of the parameters, maxItems and startItem should refer to the lowest TOC level of the description. Similarly, the values of any returned parameters also refer to the lowest level of the description. Note also that the Index Descriptor children of the lowest TOC level can also be included in the returned XML as shown below an Example J.
10EXAMPLE J Returned XML Description Fragment 1. <mageLibrary 2. requestID = `19999125` 3. noItemsIdentified = `500` 4. startItemReturned = `1` 5. noItemsRetumed = `100` 6. <Subcategory 7. textIdentifier = `Basketball[ 8. xlink:href = "http://mySite/myMetadataSvr?Category[@textIdentifier= `Sports`]/Subcategory[@textIdentifier =`Basketball`]/Image" 9. xlink:role = `description` 10. xlink:show = `embed`> 11. </Image 12. textIdentifier = `Image1` 13. xlink:href = "http://mySite/images/image1.jpg`> 14. <ImageID> Image001
</ImageID> 15. Etc. 16. </Image 17. textIdentifier = `Image2` 18. xlink:href = "http://mySite/images/i- mage2.jpg`> 19. <ImageID>Image002 </ImageID> 20. Etc. m. </Subcategory> m+1. <Subcategory m+2. textIdentifier = `Football` m+3. xlink:href = "http://mySite/myMetadataSvr?Category[@textIdentifier `Sports`]/Subcategory[@textIdentifler =`Football]/Image" m+4. xlink:role = `description` m+5. xlink:show = `embed`> m+x. Etc. n. Subcategory> n+1. </imageLibrary>
[0164] Searching Requests
[0165] Searching requests can originate from either the user specifying a structured query, using an advanced search option, or the user specifying an unstructured query, using the simple search option. The term "structured" query is used herein to mean a query that is composed of a set of constraints which are expressed in terms of known properties of the information source. These constraints can be combined in a conjunctive (and) or disjunctive (or) manner, or a mixture of the two. The term "unstructured" query is used herein to mean a query which comprises a list of keywords and expressions, with or without conjunctions (eg. the type of query used by most search engines such as Yahoo! .RTM., AltaVista.RTM., etc). The major difference between the two query types is that a structured query is formulated with a knowledge of the information source (eg. the schema of the metadata database).
[0166] If a structured query is formed then this query is preferably represented using an XPathExpression, with constraints being represented as predicates for location steps, as described for browsing in the previous section. Therefore only unstructured queries need to be considered in this section.
[0167] Most metadata collections presently in existence have an unstructured search function. In many cases considerable effort has been expended to make this search function as optimal, in terms of speed and suitable results, as possible. Consequently it is advantageous to use these search facilities whenever an unstructured query is specified by a user.
[0168] Unstructured queries can be passed to the metadata server 212 using the query function call defined earlier in this section. This function call is preferably included within a predicate of a step of a location path. As location paths can contain a predicate for each of their location steps, an XPathExpression can contain more than one unstructured query expressions. However, most requests based on unstructured queries contain a single query expression. For example, the XPathExpression, /image[query("dog OR cat")], will select all the image items that are descendents of the root node and satisfy the query "dog OR cat". Note that the XPathExpression would need to be appropriately encoded before being dispatched as part of a URI (see RFC 2396). For example, space characters should be encoded using the character triplet %20.
[0169] Typically searches can result in a large number of items. The description that is returned to the media browser 101 can be limited in the number of items the description contains by using the maxItems parameter. After receiving the first set of results, the media browser 101 may then request a further set by using the startItem parameter. To do this, the media browser 101 includes the requestID that was returned by the metadata server 212 with the response to the original request. In other words, the returned requestID identifies the start of a transaction that can be accessed by later requests.
[0170] The above has a number of implications for the configuration of the metadata server 212 as such requires the metadata server 212 to be able to save and access the results of previous requests. However, traditional server arrangements cannot maintain such results of requests in a cache indefinitely. Preferably, if a request arrives referring to a previous request, the metadata server 212 attempts to match the requestID to its cached request results. If the request is no longer in the cache, it is reprocessed. In an alternative arrangement, if a match cannot be achieved, then the metadata server 212 can optionally attempt to match the request on a textual similarity basis with other requests before resorting to reprocessing the request. This approach is helpful in that it can eliminate much duplicated processing by the metadata server 212. The size of the cache for a metadata server 212 can therefore be implementation dependent.
[0171] IV. The Media Browser Application
[0172] The media browser 101 provides the user with a single user interface for browsing and searching different metadata collections, An example graphical user interface 400 for the media browser 101 is shown in FIG. 4. The media browser interface 400 provides the user with the options of either browsing or searching for (particular items of) content via metadata associated with the (items of) content. The media browser 101 can be implemented as a stand-alone application (eg like Word97
manufactured by Microsoft Corporation of the USA) or as a service able to be supplied to multiple concurrent users. The preferred arrangement implements the media browser 101 as a service. In this mode each user is required to log in to the service to access their personalised TOC. The service aspect of the media browser 101 is discussed further in Section V below The present section is devoted to describing the functionality of the media browser 101. The description assumes a media browser service, however it should be evident that the functionality could equally well be implemented as a stand-alone program.
[0173] Typically the media browser 101 is implemented with a set of default media tool plug-ins. A user of the media browser 101 can then select, and preferably download via the Internet, further media tools to plug-in to their own implementation. Each plug-in has a defined set of target media types. The separation of media playing/viewing from metadata browsing and searching is an important concept for the media browser 101
as such allows the application to be adapted to particular users/environments.
[0174] The media browser 101 enables browsing access to the metadata by providing a Table of Contents (TOC) which represents the structure of the information landscape a user chooses to access. This information landscape can comprise links to local metadata and/or links to remote metadata and is typically customised by each user as the user discovers metadata sites which are relevant to personal interests. A default TOC is preferably provided for each new user.
[0175] The underlying information landscape is represented at all levels as a description (ie. an XML document). This means that the base structure of the description, which for XML is a tree containing nodes and links, is the same whether the user is viewing the entry point of the TOC or viewing the details of a description of a multimedia item of content (eg. a digital video). Since the TOC is a visual representation of the information landscape, the user's navigation in the TOC is unchanged for all levels of the TOC. This means that the interface 400
operates the same whether a user is browsing metadata at different Web sites, different sections of a metadata collection (eg, categories in an image nietadata collection), or within a description of multimedia content (eg. a clip in the digital video tape).
[0176] The TOC is formed by items that are selectable. These items comprise the visual representations of TOC descriptors (see Section II for more details on metadata representation). The items contain visual identifiers to aid the user in browsing. Typically, the visual identifier represents the content in some way. This is especially true for visual identifiers that correspond to items of multimedia content. Examples of visual identifiers include, simple or graphically-designed text, thumbnails of images, animations, and short previews of videos. Preferably, these visual identifiers are provided by the descriptions but, if not, the media browser 101 can graphically generate them from information contained in the description (eg. textidentifier attribute or element name). Visual identifiers have been discussed in more detail in Sections II and III.
[0177] The browsing functionality provided in the preferred arrangement may now be described with reference to FIG. 5. On activating the media browser 101, the initial description of the information landscape is read in step 500. This initial description usually contains a set of top-level links to different metadata collections or sections of metadata collections. The media browser 101 then, in step 501, processes this description and constructs an initial TOC from the description. Typically the processing of a description involves parsing the XML document that contains the description and representing the description using an object model in the computer memory. Preferably, step 501 involves detecting all the TOC descriptors from the description and building a TOC from those descriptors. Preferably, the differentiation between TOC and index descriptors is performed using the core descriptorType attribute as described in Section II.
[0178] In step 502 which follows, a view of the initial TOC is generated and presented to the user. This view may be provided in the form of a tree structure as used by applications such as WINDOWS EXPLORER manufactured by Microsoft Corporation. Preferably, a rectangular panel 402 is provided as seen in FIG. 4 showing the visual identifiers 404 that correspond to the items of the initial level of the information landscape. For example, this could be a grid of visual identifiers identifying a number of initial metadata collections.
[0179] The media browser 101 then awaits a user event. When a user selects an item, for example by clicking on a visual identifier 404 in step 503, the corresponding descriptor is examined in step 504 to determine whether it has child TOC descriptors. This may be the case if an individual description contains more than one level of structure (eg. a description of a collection may often contain several TOC levels in one description). If the item has child TOC items, then control proceeds to step 510 and the view of the TOC is updated with the child items.
[0180] If the selected item has no child TOC items, then in the step 505, the media browser 101 determines whether the item contains a link to a description. This can be achieved explicitly if the linking element, that represents the source of the link, has a specified role of "description" (roles of linking elements are described earlier). If the role of the link is undefined, then the media brow