Apparatus, Methods, and Computer Program Products For Adaptive Multimedia Content Indexing

• US 20140032538A1
• Filed: 11/29/2012
• Published: 01/30/2014
• Est. Priority Date: 07/26/2012
• Status: Active Grant

• Current Assignee: Telefonaktiebolaget LM Ericsson
• Original Assignee: Telefonaktiebolaget LM Ericsson
• Inventors: ARNGREN, Tommy, KAMAL, Arif
• Granted Patent: US 9,292,552 B2
• Time in Patent Office: Days
• US Class Current: 707/723
• CPC Class Codes: G06F 16/2228 Indexing structures
• CPC Class Codes: G06F 16/40 of multimedia data, e.g. slideshows comprising image and additional audio data retrieval of still image data G06F16/50; retrieval of audio data G06F16/60; retrieval of video data G06F16/70
• CPC Class Codes: G06F 16/41 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/435 Filtering based on additional data, e.g. user or group profiles
• CPC Class Codes: G06F 16/7834 using audio features

Abstract

This disclosure provides, among other things, a method for adaptive indexing of multimedia content. In some embodiments, the method includes: identifying a keyword associated with a feature found in a data portion of a multimedia item, the keyword being associated with a media type, and the media type being associated with a weight; determining a relevance for the keyword using the weight associated with the media type associated with the keyword; updating the weight associated with the media type based on a particular set of content items, wherein each content item included in the particular set was added to the particular set of content items because the content item was selected by a user from a set of content items that were identified to the user in response to a query from the user, thereby producing an updated weight; and updating the relevance of the keyword using the updated weight.

Claims

1. A method for adaptive indexing of multimedia content, the method comprising:

identifying a keyword associated with a feature found in a data portion of a multimedia item, the feature being of a first media type, and the first media type being associated with a first weight;

determining a relevance for the keyword using said first weight;

updating the first weight associated with the first media type based on a search result selected by a user from a set of search results that were identified to the user in response to a query from the user, thereby producing an updated first weight, wherein the selected search result indicates a multimedia item having a feature being of the first media type and being associated with the keyword; and

updating the relevance of the keyword using the updated first weight.

2. The method of claim 1, wherein the first media type is one of image, sound, voice, and text.

3. The method of claim 1, wherein the first weight associated with the first media type is associated with a set of one or more users or devices and not with any other set of users or devices.

4. The method of claim 1, wherein

the method further comprises prior to updating the first weight, determining a number of content items that have been selected by one or more users in response to one or more queries by the one or more users;

and

the step of updating the first weight occurs in response to a determination that the determined number meets or exceeds a predetermined threshold.

5. The method of claim 1, further comprising:

transmitting, from a server to a client device in response to receiving from the client device a query comprising a query keyword, a search result identifying a set of content items that match the query;

identifying a content item selected by a user of the client device from the set of items;

determining whether data associated with the selected content item indicates that the content item contains a feature that is (i) mapped to the keyword and (ii) of the same media type as the least one media type;

in response to determining that the data associated with the selected content item indicates that the content item contains a feature that is (i) mapped to the keyword and (ii) of the same media type as the least one media type, updating a counter, wherein

the step of updating the first weight comprises updating the first weight using the counter.

6. The method of claim 1, wherein the step of updating the first weight comprises updating the first weight based on a distribution of determined feature types across N queries, where N>

1.

7. The method of claim 6, wherein the updated weight is based on a ratio among current weights and a ratio among the distribution of determined feature types across the N queries.

8. The method of claim 6, wherein the updated weight is based on a weighted average of a ratio among current weights and a ratio among the distribution of determined feature types across the N queries.

9. The method of claim 1, wherein

the keyword is further associated with a second media type, and the second media type is associated with a second weight,

the step of determining the relevance for the keyword comprises calculating a relevance value using the first weight and the second weight,

the method further comprises updating the second weight based on the particular set of content items, thereby producing an updated second weight, and

the step of updating the relevance of the keyword comprises calculating a relevance value using the updated first weight and updated second weight.

10. The method of claim 1, wherein

the multimedia item is a web page comprising a first file and a second file, wherein the first file is an HTML document and the second file is one of an image file and a video file,

the keyword is further associated with a second feature of a second media type,

the first feature with which the keyword is associated was found in a data portion of the first file, and

the second feature with which the keyword is associated was found in a data portion of the second file.

11. An apparatus for adaptive indexing of multimedia content, the apparatus being configured to:

identify a keyword associated with a feature found in a data portion of a multimedia item, the feature being of a first media type, and the first media type being associated with a first weight;

determine a relevance for the keyword using the first weight;

update the first weight associated with the first media type based on a search result selected by a user from a set of search results that were identified to the user in response to a query from the user, thereby producing an updated first weight, wherein the selected search result indicates a multimedia item having a feature being of the first media type and being associated with the keyword; and

update the relevance of the keyword using the updated first weight.

12. The apparatus of claim 11, wherein the first media type is one of image, sound, voice, and text.

13. The apparatus of claim 11, wherein the first weight associated with the first media type is associated with a set of one or more users or devices and not with any other set of users or devices.

14. The apparatus of claim 11, wherein

the apparatus is further configured to determine a number of content items that have been selected by one or more users in response to one or more queries by the one or more users prior to updating the first weight, and

the apparatus is configured such that the apparatus updates the first weight in response to a determination that the determined number meets or exceeds a predetermined threshold.

15. The apparatus of claim 11, wherein the apparatus is further configured to:

transmit, in response to receiving from the client device a query comprising a query keyword, a search result identifying a set of content items that match the query;

identify a content item selected by a user of the client device from the set of items;

determine whether data associated with the selected content item indicates that the content item contains a feature that is (i) mapped to the keyword and (ii) of the same media type as the least one media type;

update a counter in response to determining that the data associated with the selected content item indicates that the content item contains a feature that is (i) mapped to the keyword and (ii) of the same media type as the least one media type; and

update the first weight using the counter.

16. The apparatus of claim 11, wherein the apparatus is configured to update the first weight based on a distribution of determined feature types across N queries, where N>

1.

17. The apparatus of claim 16, wherein the updated weight is based on a ratio among current weights and a ratio among the distribution of determined feature types across the N queries.

18. The apparatus of claim 16, wherein the updated weight is based on a weighted average of a ratio among current weights and a ratio among the distribution of determined feature types across the N queries.

19. The apparatus of claim 11, wherein

the keyword is further associated with a second media type, and the second media type is associated with a second weight,

the apparatus is configured to determine the relevance for the keyword by calculating a relevance value using the first weight and the second weight,

the apparatus is further configured to update the second weight based on the particular set of content items, thereby producing an updated second weight, and

the apparatus is configured to update the relevance of the keyword by calculating a relevance value using the updated first weight and updated second weight.

20. The apparatus of claim 11, wherein

the multimedia item is a web page comprising a first file and a second file, wherein the first file is an HTML document and the second file is one of an image file and a video file, and

said data portion of the multimedia item is a data portion of the second file.

21. A computer program product comprising a non-transitory computer readable medium storing computer instructions for adaptive indexing of multimedia content, the computer instructions comprising:

instructions for identifying a keyword associated with a feature found in a data portion of a multimedia item, the feature being of a first media type, and the first media type being associated with a first weight;

instructions for determining a relevance for the keyword using said first weight;

instructions for updating the first weight associated with the first media type based on a search result selected by a user from a set of search results that were identified to the user in response to a query from the user, thereby producing an updated first weight, wherein the selected search result indicates a multimedia item having a feature being of the first media type and being associated with the keyword; and

instructions for updating the relevance of the keyword using the updated first weight.

22. The computer program product claim 21, wherein the first media type is one of image, sound, voice, and text.

23. The computer program product claim 21, wherein the first weight associated with the first media type is associated with a set of one or more users or devices and not with any other set of users or devices.

24. The computer program product claim 21, wherein

the computer instructions further comprise instructions for determining a number of content items that have been selected by one or more users in response to one or more queries by the one or more users prior to updating the first weight;

and

the instructions for updating the first weight are configured to be executed in response to a determination that the determined number meets or exceeds a predetermined threshold.

25. The computer program product claim 21, wherein the computer instructions further comprise:

instructions for transmitting to a client device a search result identifying a set of content items that match a query;

instructions for identifying a content item selected by a user of the client device from the set of items;

instructions for determining whether data associated with the selected content item indicates that the content item contains a feature that is (i) mapped to the keyword and (ii) of the same media type as the least one media type;

instructions for updating a counter in response to a determination that the data associated with the selected content item indicates that the content item contains a feature that is (i) mapped to the keyword and (ii) of the same media type as the least one media type, wherein

the instructions for updating the first weight comprise instructions for updating the first weight using the counter.

26. The computer program product claim 21, wherein the instructions for updating the first weight comprise instructions for updating the first weight based on a distribution of determined feature types across N queries, where N>

1.

27. The computer program product claim 26, wherein the updated weight is based on a ratio among current weights and a ratio among the distribution of determined feature types across the N queries.

28. The computer program product claim 26, wherein the updated weight is based on a weighted average of a ratio among current weights and a ratio among the distribution of determined feature types across the N queries.

29. The computer program product claim 21, wherein

the keyword is further associated with a second media type, and the second media type is associated with a second weight,

the instructions for determining the relevance for the keyword comprise instructions for calculating a relevance value using the first weight and the second weight,

the computer instructions further comprise instructions for updating the second weight based on the particular set of content items, thereby producing an updated second weight, and

the instructions for updating the relevance of the keyword comprises instructions for calculating a relevance value using the updated first weight and updated second weight.

30. The computer program product claim 21, wherein

the multimedia item is a web page comprising a first file and a second file, wherein the first file is an HTML document and the second file is one of an image file and a video file,

the keyword is further associated with a second feature of a second media type,

the first feature with which the keyword is associated was found in a data portion of the first file, and

the second feature with which the keyword is associated was found in a data portion of the second file.