Estimating social interest in time-based media
Social media content items are mapped to relevant time-based media events. These mappings may be used as the basis for multiple applications, such as ranking of search results for time-based media, automatic recommendations for time-based media, prediction of audience interest for media purchasing/planning, and estimating social interest in the time-based media. Social interest in time-based media (e.g., video and audio streams and recordings) segments is estimated through a process of data ingestion and integration. The estimation process determines social interest in specific events represented as segments in time-based media, such as particular plays in a sporting event, scenes in a television show, or advertisements in an advertising block. The resulting estimates of social interest also can be graphically displayed.
1. A computer-implemented method comprising:
accessing a plurality of social media content items authored by users of a social networking system;
accessing programming information regarding an event of time based broadcast media aired during a segment of time;
annotating the event with one or more metadata instances that has semantic information about the event, the annotating using an alignment function to estimate the likelihood that the event is described by one or more of the metadata instances;
generating a set of mappings, each of the mappings generated by:
performing a feature extraction process on a pair including the annotated event and one of the social media content items, the feature extraction process converting data to a common format by generating at least one content feature that identifies co-occurring textual information between the social media content item and one of the metadata instances of the annotated event, and
generating a score for the content feature indicating whether the social media content item refers to the annotated event, using a feature specific sub-function for the content feature; and
determining a level of social interest in the event based upon the set of mappings.
2. The computer-executed method of claim 1 wherein performing the feature extraction process further comprises:
generating at least one geo-temporal feature referring to the difference in location at which the annotated event was generated from a location associated with the social media content item about the event; and
generating a score for the geo-temporal feature indicating whether the social media content item refers to the annotated event, using a feature specific sub-function for the geo-temporal feature.
3. The computer-executed method of claim 1 wherein performing the feature extraction process further comprises:
generating at least one authority feature describing information related to an author of the social media content item; and
generating a score for the authority feature indicating whether the social media content item refers to the annotated event, using a feature specific sub-function for the authority feature.
4. The computer-implemented method of claim 1 wherein the accessed social media content items have been publicly broadcasted by the social networking system.
5. The computer-implemented method of claim 1 , wherein determining a level of social interest in the event based upon the set of mappings further comprises:
determining a score by aggregating the mappings, the mappings weighted based on at least one weight determined based on the social media content item.
6. The computer-implemented method of claim 5 wherein the at least one weight comprises a social media content weight having a numerical value based on a sentiment of the social media content item.
7. The computer-implemented method of claim 5 wherein the at least one weight comprises a source-based weight having a numerical value based on a size of audience of an author of the social media content item.
8. The computer-implemented method of claim 5 wherein the at least one weight comprises a source-based weight having a numerical value based on a number of inbound links to the social media content item.
9. The computer-implemented method of claim 5 wherein the at least one weight comprises an author-based weight having a numerical value based on a demographic information of the author of the social media content item.
10. The computer-implemented method of claim 5 wherein the at least one weight comprises an author-based weight having a numerical value based on a location associated with the author of the social media content item.
11. A system comprising:
a server comprising a hardware processor and a database, the server configured to:
access a plurality of social media content items authored by users of a social networking system;
access programming information regarding an event of time based broadcast media aired during a segment of time;
annotate the event with one or more metadata instances that has semantic information about the event, the annotating using an alignment function to estimate the likelihood that the event is described by one or more of the metadata instances;
generate a set of mappings, each of the mappings generated by:
performing a feature extraction process on a pair including the annotated event and one of the social media content items, the feature extraction process converting data to a common format by generating at least one content feature that identifies co-occurring textual information between the social media content item and one of the metadata instances of the annotated event, and
generating a score for the content feature indicating whether the social media content item refers to the annotated event, using a feature specific sub-function for the content feature; and
determine a level of social interest in the event based upon the set of mappings.
12. The system of claim 11 wherein performing the feature extraction process further comprises:
generating at least one geo-temporal feature referring to the difference in location at which the annotated event was generated from a location associated with the social media content item about the event; and
generating a score for the geo-temporal feature indicating whether the social media content item refers to the annotated event, using a feature specific sub-function for the geo-temporal feature.
13. The system of claim 11 wherein performing the feature extraction process further comprises:
generating at least one authority feature describing information related to an author of the social media content item; and
generating a score for the authority feature indicating whether the social media content item refers to the annotated event, using a feature specific sub-function for the authority feature.
14. The system of claim 11 wherein the accessed social media content items have been publicly broadcasted by the social networking system.
15. The system of claim 11 , wherein determining a level of social interest in the event based upon the set of mappings further comprises:
determining a score by aggregating the mappings, the mappings weighted based on at least one weight determined based on the social media content item.
16. The system of claim 15 wherein the at least one weight comprises a social media content weight having a numerical value based on a sentiment of the social media content item.
17. The system of claim 15 wherein the at least one weight comprises a source-based weight having a numerical value based on a size of audience of an author of the social media content item.
18. The system of claim 15 wherein the at least one weight comprises a source-based weight having a numerical value based on a number of inbound links to the social media content item.
19. The system of claim 15 wherein the at least one weight comprises an author-based weight having a numerical value based on a demographic information of an author of the social media content item.
20. The system of claim 15 wherein the at least one weight comprises an author-based weight having a numerical value based on a location associated with an author of the social media content item.
21. A non-transitory computer readable medium storing instructions, the instructions when executed cause a processor to:
access a plurality of social media content items authored by users of a social networking system;
access programming information regarding an event of time based broadcast media aired during a segment of time;
annotate the event with one or more metadata instances that has semantic information about the event, the annotating using an alignment function to estimate the likelihood that the event is described by one or more of the metadata instances;
generate a set of mappings, each of the mappings generated by:
performing a feature extraction process on a pair including the annotated event and one of the social media content items, the feature extraction process converting data to a common format by generating at least one content feature that identifies co-occurring textual information between the social media content item and one of the metadata instances of the annotated event, and
generating a score for the content feature indicating whether the social media content item refers to the annotated event, using a feature specific sub-function for the content feature; and
determine a level of social interest in the event based upon the set of mappings.
22. The medium of claim 21 wherein performing the feature extraction process further comprises:
generating at least one geo-temporal feature referring to the difference in location at which the annotated event was generated from a location associated with the social media content item about the event; and
generating a score for the geo-temporal feature indicating whether the social media content item refers to the annotated event, using a feature specific sub-function for the geo-temporal feature.
23. The medium of claim 21 wherein performing the feature extraction process further comprises:
generating at least one authority feature describing information related to an author of the social media content item; and
generating a score for the authority feature indicating whether the social media content item refers to the annotated event, using a feature specific sub-function for the authority feature.
24. The medium of claim 21 wherein the accessed social media content items have been publicly broadcasted by the social networking system.
25. The medium of claim 21 , wherein determining a level of social interest in the event based upon the set of mappings further comprises:
determining a score by aggregating the mappings, the mappings weighted based on at least one weight determined based on the social media content item.
26. The medium of claim 25 wherein the at least one weight comprises a social media content weight having a numerical value based on a sentiment of the social media content item.
27. The medium of claim 25 wherein the at least one weight comprises a source-based weight having a numerical value based on a size of audience of an author of the social media content item.
28. The medium of claim 25 wherein the at least one weight comprises a source-based weight having a numerical value based on a number of inbound links to the social media content item.
29. The medium of claim 25 wherein the at least one weight comprises an author-based weight having a numerical value based on a demographic information of an author of the social media content item.
30. The medium of claim 25 wherein the at least one weight comprises an author-based weight having a numerical value based on a location associated with an author of the social media content item.