KR20050047323A - A system for multimedia retrieval and intelligent service supporting the specification of tv-anytime - Google Patents

Abstract

The present invention relates to a multimedia search and intelligent service system supporting the TV-Anytime standard, and more particularly, to a two-way TV-anytime system for a personalized broadcasting service for digital multimedia.

Multimedia retrieval and intelligent service system supporting the TV-Anytime standard of the present invention includes a metadata object management unit for storing and managing metadata; A user preference engine unit which determines personalized content based on personal preference information and content metadata; A location processor to obtain a location descriptor for location information of the content; Technical features include a metadata search API for searching for content and obtaining a reference identifier, and a metadata index manager for receiving and receiving metadata from the metadata object manager to generate and store a metadata index.

Accordingly, the multimedia retrieval and intelligence service system supporting the TV-Anytime standard of the present invention provides a future interactive broadcasting environment through a system that can be stored in a user client by searching for multimedia content and extracting a location descriptor for the corresponding content. Has the effect of providing a customized service based on the search for multimedia content and preference information for the user.

Description

A system for multimedia retrieval and intelligent service supporting the specification of TV-anytime}

The present invention relates to a multimedia search and intelligent service system supporting the TV-Anytime standard, and more particularly, to a two-way TV-anytime system for a personalized broadcasting service for digital multimedia.

A general custom method in a conventional set top box is a system that provides a user with filtering of user information and input content information in a set top box with respect to multimedia content including broadcast content.

However, the conventional system has been quite limited to deal with a variety of contents by limited resources called set-top boxes. And in the case of a customized system for the Program and System Information Protocol (PSIP) information provided through the broadcasting network in the set-top box, the information of the information transmission protocol itself is very poor, and the information to compare with the user's preference information is poor. In addition, there is a problem that the content is also limited to the broadcast content.

Accordingly, the present invention is to solve the above disadvantages and problems of the prior art, for a digital multimedia that can be provided with a customized service based on the search for multimedia content and preference information for the user in an interactive broadcast environment It is an object of the present invention to provide a two-way TV-Anytime system for a personalized broadcasting service according to the individual.

The object of the present invention is a metadata object management unit for storing and managing metadata; A user preference engine unit which determines personalized content based on personal preference information and content metadata; A location processor to obtain a location descriptor for location information of the content; Meta-data retrieval API that performs a search for content and obtains a reference identifier, and a metadata index management unit for receiving and receiving metadata from the metadata object management unit to create and store a metadata index. Is achieved by a multimedia search and intelligence service system.

Details of the above object and technical configuration of the present invention and the effects thereof according to the present invention will be more clearly understood by the following detailed description with reference to the drawings showing preferred embodiments of the present invention.

1 is a schematic diagram illustrating metadata of multimedia content. Referring to FIG. 1, metadata of multimedia content includes content description metadata, entity description metadata, and user metadata.

The content description metadata is metadata for describing content information of the content itself independently of the program entity, and one or more program entities exist for one content. In this case, each program entity may have the same reference identifier ( It has a Content Referencing Identifier (CRID) value.

The entity description metadata may be used for an application such as an electronic program guide (EPG) by describing location and service information of a program as metadata associated with a specific broadcast related program entity.

The user metadata describes information on user history and user preferences, and borrows portions related to user descriptions from the multimedia description structure of MPEG-7. Metadata about user preference describes user's taste information related to consumption of content, and compares the content with metadata (content description metadata, object related metadata, etc.) to search for, select and consume desired content. Do it.

Segment metadata breaks up content into segments and then reorganizes it into segments to provide highlights, bookmarks, event-based browsing, and table of contents views for the content. Segment-related metadata describe two types of segment information and segment group information. Segments correspond to each section divided into any number of sections in the program, and various kinds of segments for browsing content are grouped into several segments. It can be seen that the configuration.

FIG. 2 is a block diagram illustrating an X-Rey system (eXpert REsolution sYstem) according to the present invention. The multimedia retrieval and intelligence service system uses the DOM (Standard of interface proposed by W3C and Neutral Interface-Tree based API) / SAX (Method for XML parsing like DOM) for processing metadata. -Provides event-based API) interface, provides content search based on metadata and search based on attributes, and provides personalized intelligent personalized service using user profile.

Using the multimedia search and intelligence service system as shown in FIG. 2, a user can directly search for content and receive a customized service that automatically receives a list of contents suitable for the user based on user metadata created during system registration. The user can obtain a location descriptor (Locator) through the reference identifier of the received content, if the acquired location descriptor is a broadcast location descriptor can be recorded for the program, the request via the URL (On- Demand) content can be obtained by downloading the corresponding content.

Location identification performs the process of matching location-independent reference information to a temporal location (e.g. scheduled broadcast time) or a spatial location (e.g. TV channel or IP address). The selection process is output and is an unambiguous identifier that can refer to a part of the content and can be referred to by one or more reference identifiers.

The location descriptor describes the location information about when and where the content can be obtained, such as the available time of the content. The location descriptor has a form of location descriptor depending on which means the PDR (Personal Digital Recorder) acquires the content. do.

3 is a block diagram showing a server of the multimedia search and intelligent service system according to the present invention. As shown in FIG. 3, the multimedia search and intelligent service system server includes a metadata object manager 100, a user preference engine 110, a metadata index manager 120, and a location handler 130. And metadata retrieval API (Application Programming Interfaces) 140.

The content provider generates content description metadata, segment metadata, and object information metadata, which are distribution information of the content, and registers the metadata with the metadata server. The metadata registered in the metadata server is stored in the metadata database through the metadata object manager 100, parsing is performed for each part, and an index for the parsed part is generated through the metadata index manager 120. . The user performs a search for content through the metadata search API 140 and obtains a reference identifier. When the user finally determines the content acquisition, the location processor 130 obtains the location descriptor. In addition, the user profile created by the user is created as user metadata and provided to the user preference engine unit 110.

The user preference engine unit 110 determines personalized content for each user based on personal preference information and content metadata stored through an information filtering module. The personalized content is provided to the user, waits for the user's reaction, and learns and changes the user's preference through feedback on the corresponding content of the user.

4 is a block diagram showing a metadata object management unit according to the present invention. Referring to FIG. 4, the metadata object manager 100 stores and manages content description metadata, segment description metadata, object description metadata, and the like. To this end, the storage manager is responsible for creating a database schema according to the metadata schema and managing metadata such as storing and deleting changes. The metadata storage manager applies a mixed storage scheme in which metadata is parsed and managed in a database while maintaining metadata as a file.

The metadata object manager includes a metadata object manager 150, a parsing module 160, a metadata instance store 170, a metadata instance manager 180, and a metadata schema generator 190. Each module Provides an integrated interface to. The metadata to be stored is communicated to the parsing module 160 via the metadata object manager 150.

The metadata object manager 150 manages and stores metadata and delivers the metadata to the parsing module 160. The parsing module 160 includes a descriptor structure of metadata according to the TV-Anytime metadata standard. As an API for extracting Java API, Java API for XML Processing (JAXP) 1.2 proposed by W3C is used as a base. In addition, the parsing module extracts the presenter unit using the DOM / SAX interface. Based on the extracted presenter, the metadata schema generator 190 creates and manages a schema that is the basis of the metadata storage manager by checking whether a schema for the presenter is generated. The metadata instance store 170 stores the extracted description schema (DS) and its contents in each table. The content is then identified through a reference identifier and location descriptor. The metadata instance manager 180 performs operations such as change, deletion, and fetch on the stored metadata. The fetch is a function provided when a third party needs some or all of metadata. If a part of the metadata, that is, the expression schema unit fetch, is provided in the stored expression schema in the expression schema unit, the corresponding part is provided. Otherwise, the corresponding part is extracted from the entire document and transmitted. In order to effectively manage metadata, a content identifier is assigned and managed in the document storage.

5 is a block diagram illustrating a user preference engine unit according to the present invention. As shown in FIG. 5, the overall structure of the user preference engine unit for processing user preferences is shown. The information on the user used in the user preference engine unit is based on a predetermined form provided by a user interface. Explicit Declaration method for directly inputting preference information of the user, and Implicit Declaration method for inputting preferences for the program using certain buttons provided by the remote controller while watching a specific program. There is a user preference generated based on the user, and second, user usage history, which is usage information indicating a record of the user's set-top box operation, is provided.

As shown in FIG. 5, the personalization information of the user is metadataized and stored through the metadata object manager 100. The user preference engine unit 110 collects and sorts user information through the user metadata collector 220 at a suitable time when the load of the system is low, and uses the user's preference and content metadata as a preferred content list generator ( Filter through 200 to generate personalized content list information suitable for a user. The content metadata loader 230 enables the metadata retrieval API 140 to retrieve the content metadata. The personalization engine agent 210 collectively manages the user metadata collector 220, the preferred content list generator 200, and the content metadata loader 230. The information of items currently collected and compared through the user preference engine unit 110 includes title, channel, genre, cast, keyword, and time expression schema. to be.

Next, referring to a process of automatically recommending content suitable for a user by using the information in the user preference engine unit 110, first, user preference information is extracted for each item from the personal preference information index, and a title, a channel, etc. from the content index. A list of recommended programs for each item is created by searching for identifiers of programs including the user preference information for each item. The union of the created recommendation program lists for each item is obtained, and a recommendation candidate program list is created. For each content identifier in the recommendation candidate set, it is determined whether there is a list of recommended programs for each item to construct a six-dimensional preference vector, and the sum of weights of the preference vectors of each recommendation candidate program is obtained. After sorting the recommendation candidate programs based on their respective weighted sums, only the upper predetermined number (N) programs are recommended to the user.

6 is a block diagram showing the reverse file linked list structure of the metadata index management unit. First, the metadata analyzed through the parsing module is stored through the metadata object manager and then transferred to the metadata index manager. In the metadata index management unit, metadata is stored in an inverted file linked list form for speed of search and convenience of management.

As shown in FIG. 6 (a), a reverse file index for a reference identifier, a content identifier, and a location descriptor is generated and managed for location resolution, and a title, a search for content metadata, as shown in FIG. Create and manage reverse file indexes for expression schema values and reference identifiers for genres, actors, keywords, etc., and reverse file indexes for user ID and personal preference information items for personalization information as shown in FIG. Create and manage.

The reference identifier of the content list provided to the user through the user preference engine unit and the reference identifier obtained through the direct search extract the location descriptor for the corresponding content through the location processor. The location processor extracts a location descriptor for the corresponding content based on the reference identifier selected by the user. The reference identifier to extract the location descriptor is given as an input of the location processor, and the location resolution handler extracts the location descriptor of the corresponding content based on the reverse file information configured in the metadata index manager. The extracted location descriptor is sent to the user.

In addition, the location processing unit implemented in the present invention extracts the location descriptor of the corresponding content through the metadata index management unit of the own system, it is possible to extract the location descriptor through the location identification system of another location through the system expansion.

Next, another embodiment of the present invention will be described. The multimedia retrieval and intelligence service system of the present invention can be used for direct retrieval from a set-top box as a client and a customized advertisement from a service server as well as a customized service from a TV-Anytime service server. In addition, targeting advertisement service is available based on user information. This is an advertisement for a man in his twenties, who can provide advertisements or commercial advertisements for content on soccer-loving people. Refers to something that can be used for delivery and the like.

Although the present invention has been shown and described with reference to preferred embodiments as described above, it is not limited to the above-described embodiments and those skilled in the art without departing from the spirit of the present invention. Various changes and modifications will be possible.

Accordingly, the multimedia retrieval and intelligence service system supporting the TV-Anytime standard of the present invention provides a future interactive broadcasting environment through a system that can be stored in a user client by searching for multimedia content and extracting a location descriptor for the corresponding content. Has the effect of providing a customized service based on the search for multimedia content and preference information for the user.

1 is a schematic diagram illustrating metadata of multimedia content.

2 is a block diagram showing a multimedia search and intelligent service system according to the present invention.

3 is a block diagram showing a server of the multimedia search and intelligent service system according to the present invention.

4 is a block diagram showing a metadata object management unit according to the present invention.

5 is a block diagram illustrating a user preference engine unit according to the present invention.

6 is a block diagram showing the reverse file linked list structure of the metadata index management unit.

<Description of the symbols for the main parts of the drawings>

100. Metadata object management unit 110. User preference engine unit

120. Metadata index management unit 130. Location processing unit

140. Metadata Search API 150. Metadata Object Manager

160. Parsing Module 170. Metadata Instance Store

180. Metadata Instance Manager 190. Metadata Schema Generator

200. Preferred Content List Generator 210. Personalization Engine Agent

220. User Metadata Collector 230. Content Metadata Loader

Claims (11)

In the system supporting the TV-Anytime standard and providing multimedia search and intelligent service, A metadata object manager for storing and managing metadata; A user preference engine unit which determines personalized content based on personal preference information and content metadata; A location processor to obtain a location descriptor for location information of the content; A metadata retrieval API for performing a search for content and obtaining a reference identifier; And Metadata index management unit for receiving and receiving metadata from the metadata object management unit to generate and store a metadata index Multimedia search and intelligent service system supporting the TV-Anytime standard, characterized in that configured to include. The method of claim 1, The metadata object management unit A metadata object manager for storing, managing, and passing metadata to parsing modules; A parsing module for extracting the presenter structure; A metadata instance store for storing the extracted expression schema and contents in a table; A metadata instance manager for changing, deleting, and fetching stored metadata; And Metadata schema generator that checks and manages whether a schema for a presenter is created Multimedia search and intelligent service system supporting the TV-Anytime standard, characterized in that configured to include. The method of claim 2, The parsing module is a multimedia search and intelligent service system supporting the TV-Anytime standard, characterized in that using the DOM / SAX interface. The method of claim 1, And the personal preference information is at least one of title, channel, genre, actor, keyword, and time information. The method of claim 1, The user preference engine unit A user metadata collector for collecting and sorting user information; A preference content list generator configured to filter user preferences and content metadata to generate personalized content list information suitable for a user; A content metadata loader that enables a metadata retrieval API to retrieve content metadata; And And a personalization engine agent configured to collectively manage the user metadata collector, the preferred content list generator, and the content metadata loader. The method of claim 1, The user preference engine unit Extracting user preference information for each item from the personal preference information index; Creating a list of recommended programs for each item by searching for identifiers of programs including the user preference information for each item from a content index; Obtaining a union of the created recommendation program list for each item and creating a recommendation candidate program list; Determining whether a list of recommended programs for each item exists for each content identifier in the recommendation candidate set and constructing a preference vector having a dimension corresponding to the number of items; Obtaining a sum of weights of the preference vectors of each recommendation candidate program; Sorting the recommendation candidate programs based on their respective weighted sums; And Recommending only the top predetermined number of programs to the user Multimedia search and intelligent service system that supports the TV-Anytime standard, characterized in that to perform an operation including a. The method of claim 1, And the location processor extracts a location descriptor for the corresponding content based on the reference identifier selected by the user. The method of claim 1, The metadata index management unit is a multimedia retrieval and intelligent service system supporting the TV-Anytime standard, characterized in that for storing the metadata in the form of a reverse file linked list. The method of claim 1, And the metadata index manager generates and manages a reverse file index for a reference identifier, a content identifier, and a location descriptor to identify a location. The method of claim 1, The metadata index management unit creates and manages an inverse file index for an expression schema value and a reference identifier of at least one of a title, genre, actor, and keyword to search for content metadata. Multimedia search and intelligence service system that supports standards. The method of claim 1, The metadata index management unit is a multimedia search and intelligent service system that supports the TV-Anytime standard, characterized in that for generating personalized information for the user ID and personal preference information items inverse file index.