Topic and time based media affinity estimation

1. A computer-executed method, comprising:

accessing a repository comprising a first time based media event and a second time based media event;

determining a first population of social media users who are aligned with the first time based media event;

determining a second population of social media users who are aligned with the second time based media event; and

determining an affinity score indicative of affinity by social media users for both the first and second time based media events, the affinity score based on an intersection of social media users in the first and second populations.

2. The computer-executed method of claim 1 , wherein determining the first population who are aligned with the first time based media event comprises:

adding a social media user to the first population responsive to determining that at least one social media content item authored by the social user is aligned to the first time based media event.

3. The computer-executed method of claim 1 , wherein determining that the at least one social media content item authored by the social user is aligned to the first time based media event comprises:

determining a confidence score indicative of a probability that the social media content items is relevant to the time based media event; and

aligning the social media content item with the time based media event based on the confidence score.

4. The computer-executed method of claim 1 , wherein determining that the at least one social media content item authored by the social user is relevant to the first time based media event comprises:

extracting event features from annotations associated with the time based media event;

extracting social media features from the social media content item; and

determining the confidence score based on a relationship between the event features and social media features.

5. The computer-executed method of claim 1 , wherein the first and second time based media events may each be at least one from the group consisting of a TV show, a movie, a video game, an advertisement, a radio broadcast, an audio recording, and a video recording.

6. The computer-executed method of claim 1 , wherein determining an affinity score comprises:

determining an intersecting population of social media users comprising the social media users that are present in both the first and second populations;

determining the affinity score based on the intersecting population.

7. The computer-executed method of claim 6 , wherein determining the affinity score comprises calculating a weighted sum over the social media users in the intersecting population.

8. The computer-executed method of claim 7 , wherein a weight is assigned to each social media user in the intersecting population, and wherein each weight is determined based on at least one from the group consisting of: a number of times the social media user has commented on either the first or second time based media event, and a probability that the social media user comments on either the first or second time based media event.

9. The computer-executed method of claim 7 , wherein determining the affinity score comprises normalizing the weighted sum by an average probability of social media users to author social media content items regarding at least one of the first and second time based media events.

10. The computer-executed method of claim 6 wherein determining the affinity score comprises inferring the affinity score based on a correlation between a plurality of other affinity scores.

11. The computer-executed method of claim 9 , wherein inferring the affinity score based on a correlation between the other affinity scores comprises:

determining a first affinity score between the first population and a third population;

determining a second affinity score between the second population and a third population;

determining a third affinity score between the third population and the fourth population

determining the affinity score between the first and second population based on the first, second, and third affinity scores.

12. The computer-executed method of claim 6 , wherein determining the affinity score comprises normalizing the affinity score based on a size of the first or second population.

13. The computer-executed method of claim 12 , wherein determining the affinity score comprises normalizing the affinity score based on at least one from the group consisting of a cardinality of the union between the first population and the second population and a product of the cardinality between the first population and the second population.

14. The computer-executed method of claim 12 , wherein determining the affinity score comprises normalizing the affinity score based on an average affinity score between the first time based media event and a plurality of other time based media events.

15. The computer-executed method of claim 6 , wherein determining the affinity score comprises determining an expected overlap count.

16. The computer-executed method of claim 1 , comprising:

accessing an estimated real world viewing audience; and

determining a real world affinity indicative of affinity by social media users for both the first and second time based media events, the real world affinity based on the affinity score and the estimated real world viewing audience.

17. The computer-executed method of claim 1 , comprising:

filtering the first population based on a filtering criteria; and

wherein the affinity score is determined based on an intersection of social media users in the filtered first population and the second population.

18. The computer-executed method of claim 17 , wherein the filtering criteria is a demographic criteria, and wherein filtering the first population based on demographic criteria comprises:

for each social media user in the first population,

accessing social media user demographic information;

comparing the demographic criteria to the demographic information; and

removing the social media user from the first population responsive to determining that the social media user's demographic information does not match the demographic criteria.

19. The computer-executed method of claim 17 , wherein the filtering criteria is a content criteria, and wherein filtering the first population based on demographic criteria comprises:

for each social media user in the first population,

extracting social media features from a plurality of social media content items authored by the social media user;

comparing the extracted features to the content criteria; and

removing the social media user from the first population responsive to determining that none of the extracted features match the content criteria.

20. The computer-executed method of claim 17 , wherein the filtering criteria is a time criteria, and wherein filtering at least one of the first and second populations based on demographic criteria comprises:

for each social media user in the first population,

extracting a time of creation from a plurality of social media content items authored by the social media user;

comparing the times of creation to the time criteria; and

removing the social media user from the first population responsive to determining that none of the times of creation match the time criteria.

21. The computer-executed method of claim 1 , comprising:

responsive to determining that the affinity score between the first and second time based media events is greater than a threshold,

sending an advertisement to client devices associated with the second population.

22. A computer-executed method, comprising:

accessing an event repository comprising a time based media event;

aggregating a population of social media users, the aggregating comprising:

accessing a content repository comprising a social media content item authored by a social media user;

determining a confidence score indicative of a probability that the social media content item is relevant to the time based media event;

adding the social media user to the population based on the confidence score;

sending an advertisement to client devices associated with the social media users in the population.

23. A computer-executed method, comprising:

accessing a content repository comprising a time based media event;

accessing a topic repository comprising a topic;

determining a first population of social media users who are aligned with the time based media event;

determining a second population of social media users who are aligned with the topic; and

determining an affinity score indicative of affinity by social media users for both the time based media event and the topic, the affinity score based on an intersection of social media users in the first and second populations.

24. The computer-executed method of claim 1 , wherein determining the second population who are aligned with the topic comprises:

adding a social media user to the second population responsive to determining that at least one social media content item authored by the social user is aligned to the topic.

25. The computer-executed method of claim 24 , wherein determining that at least one social media content item authored by the social user is aligned to the topic comprises:

determining a confidence score indicative of a probability that the social media content items is relevant to the topic; and

aligning the social media content item with the topic based on the confidence score.

26. The computer-executed method of claim 25 , wherein determining that at least one social media content item authored by the social user is relevant to the first time based media event comprises:

extracting event features from annotations associated with the time based media event;

extracting social media features from the social media content item; and

determining the confidence score based on a relationship between the event features and social media features.

27. A computer-executed method, comprising:

accessing a topic repository comprising a first topic and a second topic;

aggregating a first population of social media users, the aggregating comprising:

determining a first confidence score indicative of a probability that a first social media content item authored by a first social media user is relevant to the first topic;

adding the first social media user to the first population based on the confidence score;

aggregating a second population of social media users, the aggregating comprising:

determining a second confidence score indicative of a probability that a second social media content item authored by a second social media user is relevant to the second topic;

adding the second social media user to the second population based on the second confidence score; and

determining an affinity score indicative of affinity by social media users for both the first and second topics, the affinity score based on an intersection of social media users in the first and second populations.