GENERATING QUERY SUGGESTIONS FROM SEMANTIC RELATIONSHIPS IN CONTENT

• US 20090313237A1
• Filed: 06/13/2008
• Published: 12/17/2009
• Est. Priority Date: 06/13/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Gade, Krishna, Agrawal, Nitin
• Granted Patent: US 8,082,278 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G06F 16/36 Creation of semantic tools, e.g. ontology or thesauri

Abstract

A method for suggesting related queries to a user query using semantic relationships that are present in informational content stored in public domains. Semantic relationships between named entities are discovered and the named entities are extracted. The entities are indexed according to the relationships. When a user query is received that includes one of the entities, query suggestions are returned to the user based on indexed relationships corresponding to the entity named in the user query.

Claims

1. One or more computer-readable storage media having machine-useable instructions embodied thereon for performing a method for suggesting additional queries related to a user query, the method comprising:

retrieving content from a content server, wherein said content comprises a plurality of instances of informational content, each instance comprising one or more entities;

parsing the content to extract at least one of the one or more entities associated with each instance of content;

identifying at least one relationship between a first extracted entity associated with a first instance of content and a second extracted entity associated with at least a second instance of content; and

creating an index based on said at least one identified relationship.

2. The media of claim 1, further comprising receiving a user query from a search engine associated with the content server, wherein said user query includes the first entity.

3. The media of claim 2, further comprising providing a query suggestion to the user, wherein the query suggestion comprises the second entity.

4. The media of claim 3, wherein the query suggestion further comprises an input region such that incident to receiving user input to the input region, the query suggestion is automatically submitted to the search engine.

5. The media of claim 1, wherein the content server comprises an academic literature database.

6. The media of claim 1, wherein the content server comprises a database associated with information relating to entertainment media.

7. One or more computer-readable storage media having machine-useable instructions embodied thereon for performing a method for suggesting additional queries related to a user query for an entity associated with content available from an informational database, the method comprising:

retrieving content from a content server, wherein said content comprises a plurality of documents, each document comprising one or more entities;

creating an index that references said documents, wherein the index comprises a relationship between at least one first entity associated with a first document and at least one second entity associated with a second document;

receiving a user query from a search engine associated with the content server, wherein the user query comprises the first entity;

searching the index to determine at least one textual match between the first entity and a relationship listed in the index; and

providing a query suggestion based on the relationship in the index.

8. The media of claim 7, wherein said content server comprises an academic literature database.

9. The media of claim 8, wherein the index is created based on extracted entities from reference logs associated with the academic literature database.

10. The media of claim 9, wherein said extracted entities comprise names of authors.

11. The media of claim 9, wherein said entities comprise topics associated with scholarly articles.

12. The media of claim 7, wherein said relationship comprises a mapping from the first entity to the second entity, wherein the first entity is a different type than the second entity.

13. The media of claim 7, wherein said relationship comprises a mapping from the first entity to the second entity, wherein the first entity is the same type as the second entity.

14. One or more computer-readable storage media having machine-useable instructions embodied thereon for performing a method for suggesting additional queries related to a user query for a first entity associated with content available from an informational database by, the method comprising:

retrieving a set of reference logs from a content server, wherein each reference log comprises a plurality of entities associated with a document stored on the content server;

parsing the set of reference logs;

extracting a first entity from a first reference log;

extracting a second entity from each reference log that comprises a relationship between the first and second entities, wherein a reference log has a relationship between the first and second entities if the reference log includes the first and second entities with not more than a predetermined number of words between them;

determining a count associated with the first and second entities, wherein the count comprises a number of times the first and second entities appear within the set of reference logs with no more than a first predetermined number of words between them;

calculating an association score for the second entity, wherein the association score comprises a weighted count associated with the first and second entities; and

creating, in a searchable index of entities, an entry that corresponds to the first entity, wherein the entry comprises a mapping between the first entity and the second entity.

15. The media of claim 14, further comprising:

extracting a plurality of additional entities, wherein each of the additional entities is extracted from each reference log that comprises a relationship between the first entity and the additional entity;

determining a plurality of counts, wherein each count is associated with the first entity and one of the additional entities;

calculating an association score for each of the additional entities;

creating, in the searchable index of entities, a plurality of additional entries corresponding to the first entity, wherein each of the additional entries comprises a mapping between the first entity and one of the additional entities; and

ranking each of the entries associated with the first entity based on the corresponding association scores.

16. The media of claim 15, further comprising receiving a user query, wherein the user query includes the first entity.

17. The media of claim 16, further comprising retrieving, from the index, a set of entries corresponding to the first entity.

18. The media of claim 17, wherein the set of entities comprises a second predetermined number of entries such that each of the set of entities is ranked higher than each entity not included in the set of entities.

19. The media of claim 14, wherein the first entity is a first type and the second entity is a second type.

20. The media of claim 19, wherein the first type comprises an author name and the second type comprises a topic.