Ranking and selecting entities based on calculated reputation or influence scores

• US 8,688,701 B2
• Filed: 12/01/2009
• Issued: 04/01/2014
• Est. Priority Date: 06/01/2007
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Topsy Labs, Inc (Apple Inc.)
• Inventors: Ghosh, Rishab Aiyer, Prakash, Vipul Ved
• Primary Examiner(s): Badawi, Sherief
• Assistant Examiner(s): Nguyen, Loan T
• Application Number:
  US12/628,791
• Publication Number:
  US 20100153404A1
• Time in Patent Office: 1,582 Days
• Field of Search: None
• US Class Current: 707/736
• CPC Class Codes: G06F 16/23 Updating
• CPC Class Codes: G06F 16/245 Query processing
• CPC Class Codes: G06F 16/24575 using context
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/335 Filtering based on additional data, e.g. user or group profiles filtering in web context G06F16/9535, G06F16/9536
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9536 Search customisation based on social or collaborative filtering
• CPC Class Codes: G06F 16/9538 Presentation of query results
• CPC Class Codes: G06F 16/954 Navigation, e.g. using categorised browsing
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

Ranking and selecting entities based on calculated reputation or influence scores is provided. In some embodiments, a method includes determining whether a first entity is a subject or an object; determining whether a second entity is a subject or an object; and generating a graph, in which a subset of the graph is a subject graph of subject nodes that includes at least one or more subjects (e.g., subject entities) linked to one or more other subjects, and in which the graph includes one or more objects (e.g., object entities) each linked to one or more subjects in the subject graph. In some embodiments, the graph includes directed and undirected links. In some embodiments, the graph includes one or more objects linked to one or more objects.

Claims

1. A method, comprising:

receiving a graph, with a subset of the graph being a subject graph of subject nodes that includes at least one or more subjects linked to one or more other subjects, the graph including one or more objects each linked to one or more subjects in the subject graph, each subject being a user, each object being a content item, and each edge in the graph representing a citation and each citation representing an expression of opinion or description by a subject on an object;

determining reputation scores for one or more subjects based on connections among the subjects in the graph, wherein the reputation scores indicate the reputation of the subjects;

selecting only a subset of citations for each object from among all the citations citing each object, the content of the citations in the selected subset matching one or more of the search terms for the search query, a different subset of citations being selectable for a same object when a different search query is provided;

assigning citation scores to any subset of a plurality of objects, the citation scores indicating relevance of the objects cited by citations and are determined based at least in part on matching one or more search terms with the content of the citations of the objects by the one or more subjects, the selection scores for an object computed for each search query based on a subset of subjects citing the object, with the subjects in the subset being the subjects of previously selected subsets of citations to each object;

combining the citation scores for the objects and the reputation scores for the subjects citing the objects to calculate selection scores for the objects determined based on matching the one or more search terms with the content of the citations, the selection scores for an object computed for each search query based on a subset of subjects citing the object, with the subjects in the subset being the subjects of previously selected subsets of citations to each object, a different selection score computed for the same object when a different search query is provided; and

selecting and ranking the objects based on the selection scores of the objects, a different ranking computed for a same set or overlapping sets of objects when the search query is different.

2. The method recited in claim 1, wherein the objects include books, films, music, documents, websites, objects for sale, objects that are reviewed or recommended or cited, or any entities that are associated with a Uniform Resource Identifier (URI), wherein the subjects include entities representing authors of Internet content or users of social media services including one or more of the following:

blogs, Twitter, and reviews on Internet web sites, wherein the links are citations of objects from subjects including one or more of the following;

Tweets, blog posts, and reviews of objects on Internet web sites.

3. The method recited in claim 1, further comprising:

receiving descriptive criteria from a user.

4. The method recited in claim 1, further comprising:

receiving search terms, wherein the search terms are provided in a search query, and wherein the search terms are included in the descriptive criteria.

5. The method recited in claim 1, further comprising:

displaying a subset of top ranked objects based on the selection scores.

6. The method recited in claim 1, further comprising:

receiving search terms, wherein the search terms are provided in a search query; and

displaying a subset of top ranked objects based on the selection scores, wherein the top ranked objects based on the selection scores provide a subjective based search result.

7. The method recited in claim 1, further comprising:

determining a content score for an object based on the description criteria;

wherein the selection score is based on the citation score, the subjective reputation score, and the content score.

8. The method recited in claim 1, further comprising:

determining an expertise score for each subject citing each object based on a second descriptive criteria, wherein the selection score is based on the citation score, the subjective reputation score, and the expertise score.

9. A computer-implemented method, comprising:

receiving a graph, with a subset of the graph being a subject graph of subject nodes that includes at least one or more subjects linked to one or more other subjects, the graph including one or more objects each linked to one or more subjects in the subject graph, each subject being a user, each object being a content item, and each edge in the graph representing a citation and each citation representing an expression of opinion or description by a subject on an object;

determining reputation scores for a first dimension for one or more objects from a perspective of a first subject based on connections among the subjects in the graph, wherein the reputation scores indicate the reputation of the subjects; and

selecting only a subset of citations for each object from among all the citations citing each object, the content of the citations in the selected subset matching one or more of the search terms for the search query, a different subset of citations being selectable for a same object when a different search query is provided;

assigning citation scores to any subset of a plurality of objects, wherein the citation scores indicate relevance of the objects cited by citations and are determined based at least in part on matching of one or more search terms with the content of the citations of the objects by the one or more subjects, the selection scores for an object computed for each search query based on a subset of subjects citing the object, with the subjects in the subset being the subjects of previously selected subsets of citations to each object;

combining the citation scores for the objects and the reputation scores for the subjects citing the objects to calculate selection scores for the objects determined based on matching of the one or more search terms with the content of the citations, the selection scores for an object computed for each search query based on a subset of subjects citing the object, with the subjects in the subset being the subjects of previously selected subsets of citations to each object, a different selection score computed for the same object when a different search query is provided;

selecting and ranking the objects based on the reputation scores and matching of one or more search terms with the content of the citations citing the objects, a different ranking computed for a same set or overlapping sets of objects when the search query is different; and

displaying the ranked objects as search result to a user or providing the search result in machine readable form.

10. The method recited in claim 9, wherein the objects include books, films, music, documents, websites, objects for sale, objects that are reviewed or recommended or cited, or any entities that are associated with a Uniform Resource Identifier (URI), wherein the subjects include entities representing Internet authors or users of social media services including one or more of the following:

blogs, Twitter, and reviews on Internet web sites, wherein the links are citations of the objects from the subjects including one or more of the following;

Tweets, Blog posts, and reviews of objects on Internet web sites.

11. The method recited in claim 9, wherein the first dimension includes search terms.

12. The method recited in claim 9, further comprising:

receiving descriptive criteria.

13. The method recited in claim 9, further comprising:

receiving search terms, wherein the search terms are provided in a search query, and wherein the search terms are included in the descriptive criteria.

14. The method recited in claim 9, further comprising:

displaying a subset of top ranked objects based on the reputation scores.

15. The method recited in claim 9, further comprising:

receiving search terms, wherein the search terms are provided in a search query; and

displaying a subset of top ranked objects based on the reputation scores, wherein the top ranked objects based on the reputation scores provide a subjective based search result.

16. The method recited in claim 9, further comprising:

determining a content score for one or more objects based on the description criteria;

wherein a selection score for each object is based on the subjective reputation score and the content score; and

ranking the one or more objects based on the selection score for each object.

17. A computer-implemented method, comprising:

receiving a graph, wherein a subset of the graph is a subject graph of subject nodes that includes at least one or more subjects linked to one or more other subjects, and wherein the graph includes one or more objects, wherein each linked to one or more subjects in the subject graph, wherein each subject is a user, and each object is a content item, and wherein each edge in the graph represents a citation, wherein each citation represents an expression of opinion or description by a subject on an object;

determining influence scores for one or more subjects based on connections among the subjects in the graph, wherein the reputation scores indicate the reputation of the subjects;

selecting only a subset of citations for each object from among all the citations citing each object, where a different subset of citations being selectable for a same object when a different search query is provided;

assigning citation scores to any subset of a plurality of the objects from the perspective of any subjects, wherein the citation score indicate relevance of the objects cited by citations and are based at least in part on matching of one or more search terms with the content of the citations of the objects by the one or more subjects;

wherein the citation scores for an object are computed for each search query as they are computed based on the selected subset of citations for the object where the content of the citations in the subset matches one or more of the search terms for the search query;

combining the citation scores for the objects and the influence scores for the subjects citing those objects to calculate selection scores for the objects determined based on matching of the one or more search terms with the content of the citations;

wherein the selection scores for an object are computed for each search query as they are computed based on a subset of subjects citing the object where the subjects in the subset are comprised of the subjects of the previously selected subsets of citations to each object; and

selecting and ranking the objects based on the selection scores of the objects; and

displaying the ranked objects as search result to a user or providing the search result in machine readable form.

18. The method recited in claim 17, wherein the objects include books, films, music, documents, websites, objects for sale, objects that are reviewed or recommended or cited, or any entities that are associated with a Uniform Resource Identifier (URI), wherein the subjects include entities representing Internet authors or users of social media services including one or more of the following:

blogs, Twitter, and reviews on Internet web sites, wherein the links are citations of the objects from the subjects including one or more of the following;

Tweets, Blog posts, and reviews of objects on Internet web sites.

19. The method recited in claim 17, further comprising:

receiving descriptive criteria.

20. The method recited in claim 17, further comprising:

receiving search terms, wherein the search terms are provided in a search query, and wherein the search terms are included in the descriptive criteria.

21. The method recited in claim 17, further comprising:

displaying a subset of top ranked objects based on the selection scores, wherein the top ranked objects based on the selection scores provide an influence based search result.

22. The method recited in claim 17, further comprising:

determining a content score for a first object based on the description criteria;

wherein the selection score is based on the citation score, the influence score, and the content score.

23. The method recited in claim 17, further comprising:

determining an expertise score for each subject citing each object based on a second descriptive criteria, wherein the selection score is based on the citation score, the influence score, and the expertise score.

24. The method recited in claim 17, further comprising:

determining a content score for a first object based on a first description criteria;

wherein the selection score is based on the citation score, the influence score, and the content score; and

determining an expertise score for each subject citing each object based on a second descriptive criteria, wherein the selection score is based on the citation score, the influence score, and the expertise score, and wherein the second descriptive criteria includes the first descriptive criteria.

25. The method recited in claim 17, wherein the selection score and object ranking is a combination of the citation score for objects and the influence score for subjects citing to the objects and an expertise score for each subject citing each object based on the citations from each subject matching descriptive criteria as a relative share of all citations from the subject, and citations from all subjects matching the descriptive criteria as a relative share of citations from all subjects.

26. The method recited in claim 17, wherein the selection score and object ranking is a combination of the citation score for objects and the reputation score for subjects citing to the object and an expertise score for each subject citing each object based on the citations from each subject matching descriptive criteria as a relative share of all citations from the subject and citations from all subjects matching the descriptive criteria as a relative share of citations from all subjects.

27. The method recited in claim 17, wherein search terms include one or more of the following search criteria:

geography, language, and time.