Scoring stream items in real time

• US 9,158,775 B1
• Filed: 04/29/2011
• Issued: 10/13/2015
• Est. Priority Date: 12/18/2010
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Tomkins, Andrew, Sivakumar, Dandapani, Sung, Sangsoo, Kosslyn, Justin, Jackson, Todd, Rohe, Andre, Luo, Ya, Bunner, Andrew, Sobol, Sasha, de Alfaro, Luca
• Primary Examiner(s): HASAN, SYED HAROON
• Application Number:
  US13/098,110
• Time in Patent Office: 1,628 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 12/06 Addressing a physical block of locations, e.g. base addressing, module addressing, memory dedication G06F12/08 takes precedence
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06F 16/435 Filtering based on additional data, e.g. user or group profiles
• CPC Class Codes: G06F 16/438 Presentation of query results
• CPC Class Codes: G06F 16/487 using geographical or spatial information, e.g. location
• CPC Class Codes: G06F 16/489 using time information
• 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: G06N 5/00 Computing arrangements using knowledge-based models
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: H04L 65/60 Network streaming of media packets
• CPC Class Codes: H04L 65/762 at the source reformatting of additional data in video distribution servers H04N21/2355
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46
• CPC Class Codes: H04L 67/60 Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources admission control or resource allocation H04L47/70

Abstract

A system and method for generating a real-time stream of content from heterogeneous data sources and a real-time index. The heterogeneous data sources include search, entertainment, social activity and activity on third-party sites. A fetching unit retrieves recent content that is indexed in the real-time index according to keywords. A model generation engine generates a model based on user activities. The mixer compares candidate content items from the heterogeneous data sources and the real-time index to the model to generate scores for each item and generates a stream of content based on the scores.

Claims

1. A method for generating a stream of content for a user in real time, the method comprising:

generating a model based on at least one interest of the user and at least one prior action from a group of heterogeneous data sources;

receiving a request from the user for a real-time content stream;

determining contextual cues surrounding the request including a time of day and a geographic location of the user associated with the request;

retrieving fresh content items from the heterogeneous data sources based upon recency;

storing and indexing the fresh content items in a real-time index;

querying the heterogeneous data sources using search terms based on the real-time index and the contextual cues for a set of candidate content items;

determining interestingness of each candidate content item to the user based on social relevance and an interest match of each item to the user;

computing a first score for each candidate content item in the set using the model and based upon the interestingness of each candidate content item to the user;

computing a threshold based at least in part on an extent of an increase in popularity within a geographic area and quality of content items having a similar subject in the geographic area;

determining whether the first score for each candidate content item in the set exceeds the threshold;

generating the stream of content in real-time from the set of candidate content items responsive to the first score for each candidate content item in the set exceeding the threshold; and

adjusting the threshold based on activities including generating the stream of content.

2. The method of claim 1, wherein the fresh content items retrieved from the heterogeneous data sources based upon recency include the fresh content items made available within a predetermined amount of time.

3. The method of claim 2, wherein the predetermined amount of time is less than 15 minutes.

4. The method of claim 1, further comprising the step of removing the candidate content items older than a predetermined amount of time from the real-time index.

5. The method of claim 1, wherein the retrieved content items are stored in the real-time index such that they can be retrieved by keywords.

6. The method of claim 1, further comprising a step of distributing the stream of content in a social network.

7. The method of claim 1, wherein generating the model uses a log of activities from the heterogeneous data sources.

8. The method of claim 1, further comprising receiving an update to the model based upon an interaction of the user with a website.

9. The method of claim 8, wherein the interaction of the user with the website includes at least one of providing a user preference, a user interest, a comment, a tag and a search.

10. The method of claim 1, further comprising the step of computing a global score for at least one candidate content item.

11. The method of claim 1, wherein the candidate content items are also received from a social server.

12. The method of claim 1, wherein computing the first score for each candidate content item uses geographic information related to at least one candidate content item.

13. The method of claim 1, wherein the heterogeneous sources comprise at least one of a search history, a blog post, a news article, a news feed, a video, a map, a message, an email message, an instant message, microblogs, a text-based post, a phone call and an activity on site.

14. The method of claim 8, wherein the interaction with the website includes determining whether the user shared something in the stream of content.

15. The method of claim 1, further comprising generating a reason for the presence of each content item in the steam of content and displaying the reason alongside each content item to the user.

16. The method of claim 1, further comprising steps of:

receiving from the user an indication that one of the candidate content items in the generated stream of content is interesting; and

generating additional content items that are related.

17. An apparatus for generating a stream of content for a user in real time, the apparatus comprising:

one or more processors;

a contextual module stored on a memory and executable by the one or more processors, the contextual module for receiving a request from the user for a real-time content stream and determining contextual cues surrounding the request including a time of day and a geographic location of the user associated with the request;

a model generation engine stored on a memory and executable by the one or more processors, the model generation engine for generating a model based on at least one interest of the user and at least one prior action from a group of heterogeneous data sources, the model having an input for modifying the model and an output for providing content items;

a fetching unit stored on the memory and executable by the one or more processors, the fetching unit for retrieving fresh content items from the heterogeneous data sources based upon recency, the fetching unit coupled to the heterogeneous data sources;

a real-time index stored on the memory and executable by the one or more processors, the real-time index coupled to the fetching unit to receive the fresh content items from the fetching unit, the real-time index for storing and indexing the fresh content items; and

a mixer stored on the memory and executable by the one or more processors, the mixer coupled to the model generation engine, the heterogeneous data sources and an indexed data store, the mixer for querying the heterogeneous data sources using search terms based on the real-time index and the contextual cues for a set of candidate content items, for determining interestingness of each candidate content item to the user based on social relevance and an interest match of each item to the user, for computing a first score for each candidate content item in the set using the model and based upon the interestingness of each candidate content item to the user, for computing a threshold based at least in part on an extent of an increase in popularity within a geographic area and quality of content items having a similar subject in the geographic area, for determining whether the first score for each candidate content item in the set exceeds the threshold, for generating the stream of content for the user in real time from the set of candidate content items responsive to the first score for each candidate content item in the set exceeding the threshold, and for adjusting the threshold based on activities including generating the stream of content.

18. The apparatus of claim 17, wherein the fetching unit also retrieves the fresh content items based on a user preference for the fresh content items from at least one of a time period, a data source, the geographic location and a list of interests.

19. The apparatus of claim 17, wherein the mixer further comprises:

a final scorer for ranking the candidate content items based upon freshness and the interestingness to produce the stream of content in real time, the final scorer coupled to the indexed data store for receiving the candidate content items.

20. The apparatus of claim 18, wherein the fetching unit retrieves the fresh content items that are made available within a predetermined amount of time.

21. The apparatus of claim 20, wherein the predetermined amount of time is less than 15 minutes.

22. The apparatus of claim 18, wherein the real-time index discards the candidate content items that are older than a predetermined amount of time.

23. The apparatus of claim 18, wherein the real-time index indexes the content items according to keywords.

24. The apparatus of claim 18, further comprising a logging unit for generating a log of activities from the heterogeneous data sources.

25. The apparatus of claim 18, wherein the mixer further comprises a global scorer for generating a global score for at least one candidate content item.

26. The apparatus of claim 18, further comprising a collaborative filter engine for transmitting the candidate content items from a social network, the collaborative filter engine coupled to the mixer.

27. The apparatus of claim 18, further comprising a user interface engine for displaying the stream of content in real time, the user interface engine coupled to the mixer.

28. A computer program product comprising a non-transitory computer useable medium including a computer readable program, wherein the computer readable program when executed on a computer causes the computer to:

generate a model based on at least one interest of a user and at least one prior action from a group of heterogeneous data sources;

receive a request from the user for a real-time content stream;

determine contextual cues surrounding the request including a time of day and a geographic location of the user associated with the request;

retrieve fresh content items from the heterogeneous data sources based upon recency;

store and index the fresh content items in a real-time index;

query the heterogeneous data sources using search terms based on the real-time index and the contextual cues for a set of candidate content items;

determine interestingness of each candidate content item to the user based on social relevance and an interest match of each item to the user;

compute a score for each candidate content item in the set using the model and based upon the interestingness of each candidate content item to the user;

compute a threshold based at least in part on an extent of an increase in popularity within a geographic area and quality of content items having a similar subject in the geographic area;

determine whether the score for each candidate content item in the set exceeds the threshold;

generate a stream of content in real-time from the set of candidate content items responsive to the score for each candidate content item in the set exceeding the threshold; and

adjust the threshold based on activities including generating the stream of content.