Identification of new sources for topics

• US 8,949,228 B2
• Filed: 01/15/2013
• Issued: 02/03/2015
• Est. Priority Date: 01/15/2013
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Hajaj, Nissan, Yaun, Garrett
• Primary Examiner(s): LE, DEBBIE M
• Application Number:
  US13/742,109
• Publication Number:
  US 20140201199A1
• Time in Patent Office: 749 Days
• Field of Search: 707/723
• US Class Current: 707/723
• 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

Abstract

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for collecting user interaction data of a plurality of users for each of a first plurality of document-text pairs, wherein the user interaction data is collected for the document-text pair from a respective service for which the respective text of the document-text pair was selected. A respective weight is calculated for each of the first plurality of document-text pairs based on, at least, the collected user interaction data for the document-text pair. One or more topics are associated with one or more of the sources based on, at least, the respective weights associated with a plurality of first document-text pairs that are associated with the source.

Claims

1. A method implemented by data processing apparatus, the method comprising:

associating a respective text with each of a plurality of documents wherein the respective text was selected based on, at least, a respective service through which the document was identified, and wherein the document and the respective text form a document-text pair;

collecting user interaction data of a plurality of users for each of a first plurality of the document-text pairs, wherein the user interaction data is collected for the document-text pair from a respective service for which the respective text of the document-text pair is selected;

calculating a respective weight for each of the first plurality of document-text pairs based on, at least, the collected user interaction data for the document-text pair; and

associating one or more topics with a source based on, at least, the respective weights associated with a plurality of first document-text pairs that associated respective documents that were published by the source, wherein the source is an author or location on a network from which content can be obtained.

2. The method of claim 1 wherein the respective text of a particular document-text pair is a query for which one or more search results responsive to the query identified the particular document of the document-text pair.

3. The method of claim 2 wherein the collected user interaction data for the particular document-text pair is selections of the search results.

4. The method of claim 1 wherein the respective text of a particular document-text pair is one or more terms from the respective document of the document-text pair.

5. The method of claim 4 wherein the collected user interaction data for the particular document text pair is selections of social network posts, electronic email, or news feed posts that identify the particular document of the particular document-text pair.

6. The method of claim 5 wherein the one or more terms are high inverse document frequency or high term frequency and inverse document frequency terms in the respective document of the particular document-text pair.

7. The method of claim 1 wherein each respective service is a search engine, a social network, electronic mail system, or a news feed system.

8. The method of claim 1 wherein each topic refers to a person, a place, a thing, or a concept.

9. The method of claim 1, further comprising:

obtaining a plurality of search results responsive to a query, each of the search results identifying a respective document published by a respective source and having a respective score; and

for one or more of the search results, determining that the respective document identified by the search result pertains to a topic associated with the respective source of respective document and adjusting the respective score of the respective document in response to the determining.

10. A system comprising:

data processing apparatus programmed to perform operations comprising;

associating a respective text with each of a plurality of documents wherein the respective text was selected based on, at least, a respective service through which the document was identified, and wherein the document and the respective text form a document-text pair;

collecting user interaction data of a plurality of users for each of a first plurality of the document-text pairs, wherein the user interaction data is collected for the document-text pair from a respective service for which the respective text of the document-text pair is selected;

calculating a respective weight for each of the first plurality of document-text pairs based on, at least, the collected user interaction data for the document-text pair; and

associating one or more topics with a source based on, at least, the respective weights associated with a plurality of first document-text pairs that associated respective documents that were published by the source, wherein the source is an author or location on a network from which content can be obtained.

11. The system of claim 10 wherein the respective text of a particular document-text pair is a query for which one or more search results responsive to the query identified the particular document of the document-text pair.

12. The system of claim 11 wherein the collected user interaction data for the particular document-text pair is selections of the search results.

13. The system of claim 10 wherein the respective text of a particular document-text pair is one or more terms from the respective document of the document-text pair.

14. The system of claim 13 wherein the collected user interaction data for the particular document text pair is selections of social network posts, electronic email, or news feed posts that identify the particular document of the particular document-text pair.

15. The system of claim 13 wherein the one or more terms are high inverse document frequency or high term frequency and inverse document frequency terms in the respective document of the particular document-text pair.

16. The system of claim 10 wherein each respective service is a search engine, a social network, electronic mail system, or a news feed system.

17. The system of claim 10 wherein each topic refers to a person, a place, a thing, or a concept.

18. The system of claim 10, wherein the operations further comprise:

obtaining a plurality of search results responsive to a query, each of the search results identifying a respective document published by a respective source and having a respective score; and

for one or more of the search results, determining that the respective document identified by the search result pertains to a topic associated with the respective source of respective document and adjusting the respective score of the respective document in response to the determining.

19. A computer program product stored on a computer readable medium that, when executed by data processing apparatus, cause the data processing apparatus to perform operations comprising:

associating a respective text with each of a plurality of documents wherein the respective text was selected based on, at least, a respective service through which the document was identified, and wherein the document and the respective text form a document-text pair;

collecting user interaction data of a plurality of users for each of a first plurality of the document-text pairs, wherein the user interaction data is collected for the document-text pair from a respective service for which the respective text of the document-text pair is selected;

calculating a respective weight for each of the first plurality of document-text pairs based on, at least, the collected user interaction data for the document-text pair; and

associating one or more topics with a source based on, at least, the respective weights associated with a plurality of first document-text pairs that associated respective documents that were published by the source, wherein the source is an author or location on a network from which content can be obtained.