Method and Apparatus for Semantic Assisted Rating of Multimedia Content

• US 20090234831A1
• Filed: 03/11/2008
• Published: 09/17/2009
• Est. Priority Date: 03/11/2008
• Status: Active Grant

• Current Assignee: SINOEAST CONCEPT LIMITED (Tencent Holdings Limited)
• Original Assignee: International Business Machines Corporation
• Inventors: Naphade, Milind, Xie, Lexing, Ebadollahi, Shahram, Smith, John Richard, Tesic, Jelena, Natsev, Apostol Ivanov, Stanoi, Ioana Roxana
• Granted Patent: US 8,032,539 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/48 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0631 Item recommendations

Abstract

The present invention is directed to a method and apparatus for assisting in rating and filtering multimedia content, such as images, videos and sound recordings. One embodiment comprises a computer implemented method for rating the objectionability of specified digital content that comprises one or more discrete content items, wherein the method includes the step of moving the specified content to one or more filtering stages in a succession of filtering stages. After the specified content is moved to a given one of the filtering stages, a rating procedure is carried out to determine whether a rating can be applied to one or more of the content items, and if so, a selected rating is applied to each of the one or more content items. The method further comprises moving content items of the specified content to the next stage in the succession after the given stage, when at least one content item of the specified content remains without rating, after the rating procedure at the given stage. When none of the content items of the specified content remains without a rating after the rating procedure has been completed at the given stage, ratings that have been respectively applied to at least some of the content items are selectively processed, in order to determine an overall objectionability rating for the specified content.

Claims

1. A computer implemented method for rating the objectionability of specified digital content comprising one or more discrete content items, wherein said method comprises the steps of:

moving said specified content to one or more filtering stages in a succession of filtering stages, the last stage of said succession being the final stage thereof;

after moving said specified content to a given one of said filtering stages, carrying out a rating procedure to determine whether a rating can be applied to one or more of said content items, and if so, applying a selected rating to each of said one or more content items;

moving content items of said specified content to the next stage in said succession after said given stage, when at least one content item of said specified content remains without a rating after said rating procedure has been completed at said given stage; and

selectively processing the ratings respectively applied to at least some of said content items, in order to determine an over-all objectionability rating for at least a portion of said specified content, when none of said content items of said specified content remains without a rating, after said rating procedure has been completed at said given stage.

2. The method of claim 1, wherein:

a given content item is rated at one of said stages by extracting semantic information therefrom, and said extracted information is used to rate the objectionability of said given content item.

3. The method of claim 1, wherein:

when a given content item is moved to said final stage of said succession, said given content item is classified into a given one of a plurality of semantic categories, wherein each category has a predetermined objectionability rating and an associated semantic model, and the rating of said given category is assigned to said given content item.

4. The method of claim 3, wherein:

a number of exemplars are used to construct each of said semantic models.

5. The method of claim 3, wherein:

each of said semantic models is constructed using a technique selected from a set of automated techniques, wherein said set includes at least use of statistical models, or use of machine learning models.

6. The method of claim 1, wherein:

each of said content items is rated with respect to a plurality of semantic dimensions, wherein said semantic dimensions include categories respectively pertaining to controversial subject matter, to non-controversial subject matter, and to correlated subject matter.

7. The method of claim 1, wherein:

semantic information is extracted from said content items for rating said content items at respective stages, wherein the information extracted for rating the content item at at least one of said stages comprises metadata associated with the content item being rated, and the information for rating a content item at least another of said stages comprises content information associated with the content item being rated.

8. The method of claim 7, wherein:

said extracted information comprises multimedia information that can include at least visual, audio and textual information.

9. The method of claim 7, wherein:

said extracted information comprises multimedia information that is associated with features which include at least color, texture, shape, skin color and graphical appearance.

10. The method of claim 1, wherein:

a given content item is rated at one of said stages in said succession by identifying the source of the given content item, and then rating the given content item on the basis of said identified source.

11. The method of claim 1, wherein:

a given content item is rated at a first stage, by determining that said given content item is identical to a first reference item having a known objectionability rating.

12. The method of claim 11, wherein:

said given content item is rated at a second stage, by determining that said given content item is nearly identical to a second reference item, to within a pre-specified limit, wherein said second reference item has a known objectionability rating.

13. The method of claim 1, wherein:

said overall objectionability rating is made available for review by a human entity.

14. The method of claim 1, wherein:

said overall objectionability rating is adaptable, selectively, for use by a particular user, or for use in a particular geographical location.

15. A computer program product in a computer readable medium for rating the objectionability of specified digital content comprising one or more discrete content items, said computer program product comprising:

instructions for moving said specified content to one or more filtering stages in a succession of filtering stages, the last stage of said succession being the final stage thereof;

instructions responsive to moving said specified content to a given one of said filtering stages, for carrying out a rating procedure to determine whether a rating can be applied to one or more of said content items, and if so, applying a selected rating to each of said one or more content items;

instructions for moving content items of said specified content to the next stage in said succession after said given stage, when at least one content item of said specified content remains without a rating after said rating procedure has been completed at said given stage; and

instructions for selectively processing the ratings respectively applied to at least some of said content items, in order to determine an over-all objectionability rating for at least a portion of said specified content, when none of said content items of said specified content remains without a rating, after said rating procedure has been completed at said given stage.

16. The computer program product of claim 15, wherein:

when a given content item is moved to said final stage of said succession, said given content item is classified into a given one of a plurality of semantic categories, wherein each category has a predetermined objectionability rating and an associated semantic model, and the rating of said given category is assigned to said given content item.

17. The computer program product of claim 15, wherein:

semantic information is extracted from said content items for rating said content items at respective stages, wherein the information extracted for rating the content item at at least one of said stages comprises metadata associated with the content item being rated, and the information for rating a content item at least another of said stages comprises content information associated with the content item being rated.

18. Apparatus for rating the objectionability of specified digital content comprising one or more discrete content items, wherein said apparatus comprises:

means for moving said specified content to one or more filtering stages in a succession of filtering stages, the last stage of said succession being the final stage thereof;

means responsive to moving said specified to a given one of said filtering stages, for carrying out a rating procedure to determine whether a rating can be applied to one or more of said content items, and if so, applying a selected rating to each of said one or more content items;

means for moving content items of said specified content to the next stage in said succession after said given stage, when at least one content item of said specified content remains without a rating after said rating procedure has been completed at said given stage; and

means for selectively processing the ratings respectively applied to at least some of said content items, in order to determine an over-all objectionability rating for at least a portion of said specified content, when none of said content items of said specified content remains without a rating, after said rating procedure has been completed at said given stage.

19. The apparatus of claim 18, wherein:

said apparatus includes a module for classifying a given content item at said final stage of said succession into a given one of a plurality of semantic categories, wherein each category has a predetermined objectionability rating and an associated semantic model, and the rating of said given category is assigned to said given content item.

20. The apparatus of claim 19, wherein:

each of said semantic models is constructed using a technique selected from a set of automated techniques, wherein said set includes at least use of statistical models, and use of machine learning models.