Blending by query classification on online social networks

• US 9,754,037 B2
• Filed: 08/27/2014
• Issued: 09/05/2017
• Est. Priority Date: 08/27/2014
• Status: Active Grant

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Inventors: Ayan, Necip Fazil, Boucher, Maxime, Li, Xiao, Perelygin, Alexander
• Primary Examiner(s): Jami, Hares
• Application Number:
  US14/470,583
• Publication Number:
  US 20160063115A1
• Time in Patent Office: 1,105 Days
• Field of Search: 707722, 707758, 707769, 707771
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

In one embodiment, a method includes receiving a search query from a first user and identifying one or more second nodes that match the search query. The method includes determining one or more search intents of the search query. Search intent may be based on one or more topics associated with the identified nodes and one or more node-types of the identified nodes. The method includes generating one or more search results corresponding to the search query, the search-results being generated based on the determined search intents. The method includes sending a search-results page to the client system of the first user for display. The search-results page may include one or more of the generated search results.

Claims

1. A method comprising:

accessing a social graph comprising a plurality of nodes and a plurality of edges connecting the nodes, the nodes comprising;

a first node corresponding to a first user associated with an online social network; and

a plurality of second nodes corresponding to a plurality of objects associated with the online social network, respectively, each object being of a particular object-type, each second node being of a node-type corresponding to the object-type of the respective corresponding object;

receiving from a client system of the first user a search query;

identifying one or more second nodes that match the search query;

determining one or more search intents of the search query, the determined search intents being based on one or more topics associated with the identified second nodes and a ratio of one or more node-types of the identified second nodes that match the search query, wherein the determined search intent indicates a query type of the search query;

generating one or more search-result modules corresponding to the search query, the search-result modules being generated based on the one or more topics and one or more node-types associated with the determined search intents of the search query, each search-result module being associated with a particular object-type and comprising one or more search results, each search result comprising a reference to one of the identified second nodes having a node-type corresponding to the particular object-type of the respective search-result module; and

sending, responsive to receiving the search query, a blended search-results page to the client system of the first user for display, the blended search-results page comprising a threshold number of the generated search-result modules, the search-result modules being presented in a blended order, the ordering of each search-result module being based on the determined search intents of the search query and a distribution of the one or more topics and one or more node-types of the determined search intents with respect to the particular object-type associated with the search-result module.

2. The method of claim 1, wherein the search query is an unstructured text query, and the method further comprises parsing the text query to identify one or more n-grams, wherein at least one of the n-grams is an ambiguous n-gram;

and wherein identifying one or more second nodes that match the search query comprises identifying one or more second nodes matching the ambiguous n-gram, each identified node corresponding to one or more second nodes of the plurality of second nodes.

3. The method of claim 1, wherein identifying one or more second nodes that match the search query comprises searching a plurality of verticals to identify a plurality of sets of objects in each vertical, respectively, that match the search query, wherein each set of objects corresponds to a search-result module, and wherein each vertical stores one or more objects associated with the online social network, each object corresponding to a second node of the plurality of second nodes.

4. The method of claim 3, wherein each vertical of the plurality of verticals stores objects of a particular object-type, the particular object-type being selected from a group consisting of:

a user, a photo, a post, a page, an application, an event, a location, or a user group.

5. The method of claim 4, wherein the determined search intent is further based on the object-type associated with each vertical having at least one identified second node.

6. The method of claim 1, wherein identifying one or more second nodes that match the search query comprises identifying at least one candidate node corresponding to the search query, wherein each candidate node corresponds to an electronic topic index.

7. The method of claim 6, wherein identifying one or more second nodes that match the search query further comprises:

receiving an electronic topic index comprising a set of nodes, wherein each node represents a topic;

identifying an anchor term in the query;

identifying candidate nodes based on the anchor term, wherein candidate nodes comprise electronic-topic-index nodes representing subjects related to the anchor term;

determining a context of the anchor term in the query;

determining a score for each of the one or more of the candidate nodes based on the determined context; and

determining whether there is a candidate node to represent the meaning of the anchor term based on the determined scores.

8. The method of claim 7, wherein receiving an electronic topic index comprises:

retrieving a database of articles, wherein one or more pairs of articles are linked;

creating a node for each of the one or more of the articles, the node comprising the topic of the article; and

for each pair of nodes corresponding to linked articles, connecting the pair of nodes with an edge.

9. The method of claim 8, wherein the database of articles comprises a web-based database, wherein each article is represented by a web page within the web-based database, and wherein two articles are linked if the web page representing one of the articles comprises a URL link to the other article.

10. The method of claim 9, wherein the determined search intent is further based on the electronic-topic-index nodes associated with at least one candidate node that represents the meaning of the anchor term.

11. The method of claim 1, further comprising scoring the search results based on the one or more determined search intents.

12. The method of claim 11, wherein the search-results page comprises one or more search results having a score greater than a threshold score.

13. The method of claim 1, wherein the particular object-type associated with each search-result module is selected from a group consisting of:

a user, a photo, a post, a page, an application, an event, a location, or a user group.

14. The method of claim 1, wherein the determined search intent is further based on matching the search query to one or more grammars of a grammar model.

15. The method of claim 1, wherein the search-results page is a user interface of a native application associated with the online social network on the client system of the first user.

16. The method of claim 1, wherein the search-results page is a webpage of the online social network accessed by a browser client of the client system of the first user.

17. The method of claim 1, wherein the one or more topics associated with the identified second nodes and the one or more topics associated with the determined search intents are determined based on a set of subjects associated with an electronic topic index.

18. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:

access a social graph comprising a plurality of nodes and a plurality of edges connecting the nodes, the nodes comprising;

a first node corresponding to a first user associated with an online social network; and

a plurality of second nodes corresponding to a plurality of objects associated with the online social network, respectively, each object being of a particular object-type, each second node being of a node-type corresponding to the object-type of the respective corresponding object;

receive from a client system of the first user a search query;

identify one or more second nodes that match the search query;

determine one or more search intents of the search query, the determined search intents being based on one or more topics associated with the identified second nodes and a ratio of one or more node-types of the identified second nodes that match the search query, wherein the determined search intent indicates a query type of the search query;

generate one or more search-result modules corresponding to the search query, the search-result modules being generated based on the one or more topics and one or more node-types associated with the determined search intents of the search query, each search-result module being associated with a particular object-type and comprising one or more search results, each search result comprising a reference to one of the identified second nodes having a node-type corresponding to the particular object-type of the respective search-result module; and

send, responsive to receiving the search query, a blended search-results page to the client system of the first user for display, the blended search-results page comprising a threshold number of the generated search-result modules, the search-result modules being presented in a blended order, the ordering of each search-result module being based on the determined search intents of the search query and a distribution of the one or more topics and one or more node-types of the determined search intents with respect to the particular object-type associated with the search-result module.

19. A system comprising:

one or more processors; and

a memory coupled to the processors comprising instructions executable by the processors, the processors operable when executing the instructions to;

access a social graph comprising a plurality of nodes and a plurality of edges connecting the nodes, the nodes comprising;

a first node corresponding to a first user associated with an online social network; and

a plurality of second nodes corresponding to a plurality of objects associated with the online social network, respectively, each object being of a particular object-type, each second node being of a node-type corresponding to the object-type of the respective corresponding object;

receive from a client system of the first user a search query;

identify one or more second nodes that match the search query;

determine one or more search intents of the search query, the determined search intents being based on one or more topics associated with the identified second nodes and a ratio of one or more node-types of the identified second nodes that match the search query, wherein the determined search intent indicates a query type of the search query;

generate one or more search-result modules corresponding to the search query, the search-result modules being generated based on the one or more topics and one or more node-types associated with the determined search intents of the search query, each search-result module being associated with a particular object-type and comprising one or more search results, each search result comprising a reference to one of the identified second nodes having a node-type corresponding to the particular object-type of the respective search-result module; and

send, responsive to receiving the search query, a blended search-results page to the client system of the first user for display, the blended search-results page comprising a threshold number of the generated search-result modules, the search-result modules being presented in a blended order, the ordering of each search-result module being based on the determined search intents of the search query and a distribution of the one or more topics and one or more node-types of the determined search intents with respect to the particular object-type associated with the search-result module.