Query-independent entity importance in books

• US 7,783,644 B1
• Filed: 12/13/2006
• Issued: 08/24/2010
• Est. Priority Date: 12/13/2006
• Status: Expired due to Fees

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Petrou, David, Reynar, Jeffrey C., Chan, Chiu-Ki, Jevtic, Nikola, Loreto, Daniel
• Primary Examiner(s): Trujillo; James
• Assistant Examiner(s): Casanova; Jorge A
• Application Number:
  US11/610,446
• Time in Patent Office: 1,350 Days
• Field of Search: 707/2, 707/102, 707/999.002, 707/999.102, 707/748
• US Class Current: 707/748
• CPC Class Codes: G06F 16/3331 Query processing
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A corpus contains information including text from books and metadata about the books. The book texts mention entities of various types, such as location, date, event, and person entities. An entity importance engine analyzes the information in the corpus to identify the entities mentioned therein, and ranks the entities using query-independent importance scores. The importance scores can be based in part on the contexts in which the entities are mentioned by the books. A presentation engine enables searching of the corpus to identify books satisfying a search query. The presentation engine presents information about a designated book. Entities mentioned in the book are presented in a style selected to enhance comprehension and utility. For example, location entities can be presented on a map with markers showing the locations of the entities. The number of entities presented are limited and ordered based on the query-independent importance scores.

Claims

1. A system for presenting entities mentioned in books, comprising:

a non-transitory computer-readable storage medium storing executable computer program instructions comprising;

an entity importance engine for assigning query-independent importance scores to entities of an entity type mentioned by books in a corpus, the entity importance engine comprising a references module for identifying third party references to sections of a book in which an entity is mentioned, assigning importance values to the sections of the book, and calculating an importance score for the entity responsive at least in part to the importance values, wherein the importance values for sections of the book referenced by third parties are elevated relative to the importance values for sections of the book not referenced by third parties;

a search module for receiving a search query from a requestor and identifying a list of a plurality of books in the corpus that at least partially satisfy the query, and ranking the books in the list in an order based at least in part on the query-independent importance scores assigned to entities mentioned by the books; and

a presentation module for presenting the plurality of books to the requestor in the ranked order; and

a processor for executing the computer program instructions.

2. The system of claim 1, wherein there are a plurality of entity types, and wherein the importance scores assigned to the entities indicate relative importance of entities of a same type.

3. The system of claim 1, wherein the entity importance engine further comprises:

an importance calculation module for calculating the importance scores for the entities, the importance calculation module comprising one or more modules from the set consisting of;

a book context module for evaluating a context in which the entity is mentioned by the book and calculating the importance score for the entity responsive at least in part to the context;

a precision module for determining a precision at which the entity is mentioned in the book and calculating the importance score for the entity responsive at least in part to the precision;

a user activities module for identifying one or more sections of a book in which the entity is mentioned that are accessed more frequently than other sections and calculating the importance score for the entity responsive at least in part to a frequency of access; and

a term weighting module for determining frequencies that entities are mentioned in the corpus, determining a frequency of which the entity is mentioned in the book relative to a frequency that the entity is mentioned in other books in the corpus and calculating the importance score for the entity responsive at least in part to the frequency of mentions.

4. The system of claim 1, wherein the entity importance engine further comprises:

a ranking module for calculating final importance scores as weighted sums of individual importance scores assigned to the entities and ranking the entities in order of importance responsive to the entities'"'"' final importance scores.

5. The system of claim 1, wherein there are a plurality of entity types, and wherein the presentation module presents entities of a single type in a particular style selected based at least in part on the entity type.

6. The system of claim 5, wherein an entity type is a geographic location, and wherein the presentation module presents geographic location entities on a map.

7. The system of claim 5, wherein an entity type is a date, and wherein the presentation module presents date entities on a timeline.

8. A computer program product stored on a non-transitory computer-readable storage medium having executable computer program instructions for presenting entities mentioned in books, comprising:

an entity importance engine for assigning query-independent importance scores to entities mentioned by books in a corpus, the entity importance engine comprising a library classification module for determining whether library classification data mention an entity in a book and calculating an importance score for the entity responsive at least in part to the library classification data, wherein the importance scores of entities mentioned in the library classification data are elevated relative to importance scores of entities not mentioned in the library classification data;

a search module for receiving a search query from a requestor and identifying a list of a plurality of books in the corpus that at least partially satisfy the query, and ranking the books in the list in an order based at least in part on the query-independent importance scores assigned to entities mentioned by the books; and

a presentation module for presenting the plurality of books to the requestor in the ranked order.

9. The computer program product of claim 8, wherein there are a plurality of entity types, and wherein the importance scores assigned to the entities indicate relative importance of entities of a same type.

10. The computer program product of claim 8, wherein the entity importance engine further comprises:

an importance calculation module for calculating the importance scores for the entities, the importance calculation module comprising one or more modules from the set consisting of;

a book context module for evaluating a context in which the entity is mentioned by the book and calculating the importance score for the entity responsive at least in part to the context;

a precision module for determining a precision at which the entity is mentioned in the book and calculating the importance score for the entity responsive at least in part to the precision;

a references module for identifying third party references to sections of the book in which the entity is mentioned and calculating the importance score for the entity responsive at least in part to the third party references;

a user activities module for identifying one or more sections of a book in which the entity is mentioned that are accessed more frequently than other sections and calculating the importance score for the entity responsive at least in part to a frequency of access; and

a term weighting module for determining frequencies that entities are mentioned in the corpus, determining a frequency of which the entity is mentioned in the book relative to a frequency that the entity is mentioned in other books in the corpus and calculating the importance score for the entity responsive at least in part to the frequency of mentions.

11. The computer program product of claim 8, wherein the entity importance engine comprises:

a ranking module for calculating final importance scores as weighted sums of individual importance scores assigned to the entities and ranking the entities in order of importance responsive to the entities'"'"' final importance scores.

12. The computer program product of claim 8, wherein there are a plurality of entity types, and wherein the presentation module presents entities of a single type in a particular style selected based at least in part on the entity type.

13. The computer program product of claim 12, wherein an entity type is a geographic location, and wherein the presentation module presents geographic location entities on a map.

14. The computer program product of claim 12, wherein an entity type is a date, and wherein the presentation module presents date entities on a timeline.

15. A computer-implemented method of determining an importance score for an entity mentioned in a book, comprising:

identifying third-party references to sections of the book, the third party references external to the book;

assigning importance values to the sections of the book, wherein the importance values for sections of the book referenced by third parties are elevated relative to the importance values for sections of the book not referenced by third parties;

identifying one or more sections of the book that mention the entity; and

calculating the importance score for the entity responsive at least in part to the importance values of the one or more sections of the book that mention the entity, the importance score indicating an importance of the entity relative to other entities mentioned in the book.

16. The method of claim 15, further comprising:

calculating importance scores for a plurality of entities responsive to the importance values of sections of the book that mention the entities;

selecting a subset of the plurality of entities responsive to the entities'"'"' importance scores; and

presenting the selected subset of entities to a user.

17. The method of claim 15, wherein the importance values assigned to sections of the book indicate relative importance of the sections, and wherein entities mentioned in more important sections receive elevated importance scores.

18. The method of claim 15, further comprising:

identifying a page of the book that mention the entity;

determining a location on the page wherein the entity is mentioned; and

calculating the importance score for the entity responsive to the location on the page.

19. The method of claim 15, wherein assigning importance values to sections of the book comprises:

determining whether sections of the book reference other sections of the book; and

elevating importance values assigned to sections of the book that are referenced by other sections of the book.

20. The method of claim 15, further comprising:

identifying an entity mentioned in a third-party reference to a section of the book that is also mentioned in the referenced section of the book; and

elevating an importance score for the entity mentioned in the third-party reference and the referenced section of the book relative to entities in the referenced section of the book that are not mentioned in the third party reference.

21. The method of claim 15, wherein assigning importance values to sections of the book comprises:

identifying a section of the book that is accessed more frequently than other sections of the book; and

elevating an importance value of the more frequently-accessed section.

22. A computer-implemented method of determining an importance score for an entity mentioned in a book, comprising:

identifying library classification data indicating a classification of the book;

determining whether the library classification data mention entities from the book;

assigning query-independent importance scores to the entities of an entity type mentioned in the book, wherein the importance scores of entities mentioned in the library classification data are elevated relative to importance scores of entities not mentioned in the library classification data;

selecting a subset of the plurality of entities responsive to the entities'"'"' importance scores and entity type; and

presenting the selected subset of entities to a user.

23. The method of claim 22, wherein assigning query-independent importance scores comprises:

determining a precision at which the book mentions an entity;

calculating the importance score for the entity responsive at least in part to the precision at which the book mentions the entity, wherein an entity mentioned more precisely receives an elevated importance score relative to an entity mentioned less precisely.

24. The method of claim 22, wherein assigning query-independent importance scores comprises:

determining an optical character recognition (OCR) confidence score for book text related to an entity; and

calculating the importance score for the entity responsive at least in part to the OCR confidence score.

25. The method of claim 22, wherein the book is in a corpus comprising a plurality of books and wherein assigning query-independent importance scores comprises:

determining a frequency that an entity is mentioned in the book and a frequency that the entity is mentioned in other books in the corpus; and

calculating the importance score for the entity based at least in part on the frequency of which the entity is mentioned in the book and the frequency that the entity is mentioned in other books in the corpus.