Assumption mechanism for queries

• US 9,213,768 B1
• Filed: 05/14/2010
• Issued: 12/15/2015
• Est. Priority Date: 05/15/2009
• Status: Active Grant

• Current Assignee: Wolfram Alpha LLC (Wolfram Research Incorporated)
• Original Assignee: Wolfram Alpha LLC (Wolfram Research Incorporated)
• Inventors: Wolfram, Stephen, Tafjord, Oyvind
• Primary Examiner(s): Fan, Shiow-Jy
• Application Number:
  US12/780,685
• Time in Patent Office: 2,041 Days
• Field of Search: 707/723, 707/713, 707/736, 707/769, 707/779
• US Class Current: 1/1
• CPC Class Codes: G06F 16/245 Query processing
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G06F 16/3338 Query expansion
• CPC Class Codes: G06F 16/90324 using system suggestions
• CPC Class Codes: G06F 16/90335 Query processing
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

A word meaning or sense is determined in response to an actual user query in a system designed for computation, search, or other activities involving the processing of queries. Assumptions, which are attempts to disambiguate pieces of the user'"'"'s query string, are generated. Generation of the assumptions is based on various heuristics and algorithms which are designed to disambiguate words which may connote multiple different entities, meanings, or senses given the context. Assumptions also may be ranked based on a set of criteria that does not require an elaborately constructed body of world knowledge to be created. A most likely assumption may be chosen for generating a machine-generated query and/or for generating query results.

Claims

1. A method, comprising:

receiving, at one or more computing devices, a query that includes a word having multiple meanings, having multiple senses, or that corresponds to multiple entities;

when the word is not recognized by the one or more computing devices;

determining, based at least in part on relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, one or more entities to which the word corresponds, or

choosing, based at least in part on the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, one meaning or sense;

when the word is recognized by the one or more computing devices and the word refers to multiple entities, determining, with one or more computing devices, the entities to which the word corresponds;

when the word corresponds to multiple entities in a same category, ranking, with one or more computing devices, the multiple entities in the same category to which the word corresponds using a set of attributes common to entities in the category, wherein the set of attributes is different than other sets of attributes common to entities in other categories;

when the word is recognized by the one or more computing devices and the word has multiple meanings or senses, ranking, with one or more computing devices, the multiple meanings or senses of the word based on attributes of a user;

when the word is recognized by the one or more computing devices and the word has multiple meanings or senses, ranking, with one or more computing devices, one or more meanings or senses of the word based on a measure of popularity of the one or more meanings or senses;

when the word corresponds to multiple entities in a same category, or the word is recognized by the one or more computing devices and the word has multiple meanings or senses, choosing, with one or more computing devices, one meaning or sense based on the rankings; and

generating, with one or more computing devices, either;

a new query and query results based on the one meaning or sense, or

query results based on the one meaning or sense.

2. The method according to claim 1, wherein the new query includes information to indicate the one meaning or sense.

3. The method according to claim 2, wherein the information to indicate the one meaning or sense is a keyword that indicates the one meaning or sense.

4. The method according to claim 1, wherein the query results include information to indicate the one meaning or sense.

5. The method according to claim 1, further comprising determining one or more additional meanings or senses based on the rankings.

6. The method according to claim 5, wherein the query results include information to indicate the one or more additional meanings or senses.

7. The method according to claim 5, further comprising:

providing a user interface mechanism to permit a user to select one of the one or more additional meanings or senses;

receiving an indication of a selected additional meaning or sense selected by the user using the user interface mechanism; and

generating a new query based on the selected additional meaning or sense and/or returning query results based on the selected additional meaning or sense.

8. The method according to claim 5, further comprising generating a web page that includes:

an indication of the one meaning or sense,

the query results based on the one meaning or sense, and

indications of the one or more additional meanings or senses.

21. The method according to claim 1, further comprising:

when the word is not recognized by the one or more computing devices, generating a keyword based on an analysis of the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, wherein determining the entities to which the word corresponds or choosing one meaning or sense is based on the keyword.

9. A tangible, non-transitory storage medium having stored thereon machine executable instructions, the machine executable instructions, when executed by one or more machines, cause the one or more machines to:

receive a query that includes a word having multiple meanings, having multiple senses, or that corresponds to multiple entities;

when the word is not recognized by the one or more machines;

determine, based at least in part on relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, one or more entities to which the word corresponds, or

choose, based at least in part on the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, one meaning or sense;

when the word is recognized by the one or more machines and the word refers to multiple entities, determine the entities to which the word corresponds;

when the word corresponds to multiple entities in a same category, rank the multiple entities in the same category to which the word corresponds using a set of attributes common to entities in the category, wherein the set of attributes is different than other sets of attributes common to entities in other categories;

when the word is recognized by the one or more machines and the word has multiple meanings or senses, rank the multiple meanings or senses of the word based on attributes of a user;

when the word is recognized by the one or more computing devices and the word has multiple meanings or senses, rank one or more meanings or senses of the word based on a measure of popularity of the one or more meanings or senses;

when the word corresponds to multiple entities in a same category, or the word is recognized by the one or more machines and the word has multiple meanings or senses, choose one meaning or sense based on the rankings; and

generate either;

a new query and query results based on the one meaning or sense, or query results based on the one meaning or sense.

22. The tangible, non-transitory storage medium according to claim 9, the machine executable instructions, when executed by the one or more machines, further cause the one or more machines to:

when the word is not recognized by the one or more machines, generate a keyword based on an analysis of the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, wherein determining the entities to which the word corresponds or choosing one meaning or sense is based on the keyword.

10. A method, comprising:

transmitting, via a network to a query answering system, a query in an imprecise syntax, the query including a word having multiple meanings or senses;

receiving, via the network, query results that are based on a first meaning or sense of the word, wherein;

when the word is not recognized by the query answering system, the query answering system;

determines, based at least in part on relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, one or more entities to which the word corresponds, or

chooses, based at least in part on the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, the first meaning or sense;

when the word is recognized by the query answering system and the word refers to multiple entities, the query answering system determines, with one or more computing devices, the entities to which the word corresponds;

when the word corresponds to multiple entities in a same category, the query answering system ranks, with one or more computing devices, the multiple entities in the same category to which the word corresponds using a set of attributes common to entities in the category, wherein the set of attributes is different than other sets of attributes common to entities in other categories;

when the word is recognized by the query answering system and the word has multiple meanings or senses, the query answering system ranks, with one or more computing devices, the multiple meanings or senses of the word based on attributes of a user;

when the word is recognized by the query answering system and the word has multiple meanings or senses, the query answering system ranks, with one or more computing devices, one or more meanings or senses of the word based on a measure of popularity of the one or more meanings or senses; and

when the word corresponds to multiple entities in a same category, or the word is recognized by the query answering system and the word has multiple meanings or senses, the query answering system chooses, with one or more computing devices, the first meaning or sense based on the rankings;

receiving, via the network, an indication of the first meaning or sense of the word, wherein the first meaning or sense corresponds to an entity in a category;

receiving, via the network, a user interface mechanism to permit selection of another meaning or sense from a set of one or more meanings or senses different than the first meaning or sense; and

displaying the query results, the indication of the first meaning or sense of the word, and the user interface mechanism to permit selection of another meaning or sense on a display device, wherein a list of entities in the same category to which the word corresponds are displayed in an order according to a set of attributes common to entities in the category, wherein the set of attributes is different than other sets of attributes common to entities in other categories.

11. The method according to claim 10, further comprising:

receiving, via the network, a user interface mechanism to request further query results based on the selected another meaning or sense; and

displaying the user interface mechanism user interface mechanism to request a further query results on the display device.

12. The method according to claim 10, wherein the user interface mechanism is for requesting further query results based on the selected another meaning or sense.

13. The method according to claim 10, further comprising:

selecting, using the user interface mechanism, another meaning or sense from the set of one or more meanings or senses different than the first meaning or sense;

transmitting, via the network, an indication of the selected another meaning or sense;

receiving further query results based on the selected another meaning or sense.

23. The method according to claim 10, wherein when the word is not recognized by the query answering system, the query results are determined from a keyword generated based on an analysis of the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word.

14. A tangible, non-transitory storage medium having stored thereon machine executable instructions, the machine executable instructions, when executed by one or more machines, cause the one or more machines to:

cause a query to be transmitted via a network to a query answering system, wherein the query is in an imprecise syntax and includes a word having multiple meanings or senses;

receive query results that are based on a first meaning or sense of the word in response to the query, wherein;

when the word is not recognized by the query answering system, the query answering system;

determines, based at least in part on relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, one or more entities to which the word corresponds, or

chooses, based at least in part on the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, the first meaning or sense;

when the word is recognized by the query answering system and the word refers to multiple entities, the query answering system determines, with one or more computing devices, the entities to which the word corresponds;

when the word corresponds to multiple entities in a same category, the query answering system ranks, with one or more computing devices, the multiple entities in the same category to which the word corresponds using a set of attributes common to entities in the category, wherein the set of attributes is different than other sets of attributes common to entities in other categories;

when the word is recognized by the query answering system and the word has multiple meanings or senses, the query answering system ranks, with one or more computing devices, the multiple meanings or senses of the word based on attributes of a user;

when the word is recognized by the query answering system and the word has multiple meanings or senses, the query answering system ranks, with one or more computing devices, one or more meanings or senses of the word based on a measure of popularity of the one or more meanings or senses; and

when the word corresponds to multiple entities in a same category, or the word is recognized by the query answering system and the word has multiple meanings or senses, the query answering system chooses, with one or more computing devices, the first meaning or sense based on the rankings;

receive an indication of the first meaning or sense of the word in response to the query, wherein the first meaning or sense corresponds to an entity in a category;

receive a user interface mechanism to permit selection of another meaning or sense from a set of one or more meanings or senses different than the first meaning or sense in response to the query; and

cause the query results, the indication of the first meaning or sense of the word, and the user interface mechanism to be displayed on a display device, wherein a list of entities in the same category to which the word corresponds are displayed in an order according to a set of attributes common to entities in the category, wherein the set of attributes is different than other sets of attributes common to entities in other categories.

24. The tangible, non-transitory storage medium according to claim 14, wherein when the word is not recognized by the query answering system, the query results are determined from a keyword generated based on an analysis of the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word.

15. A method, comprising:

receiving, at one or more computing devices, a query in an imprecise syntax, the query including a word, phrase, or group of textual characters;

when the word, phrase, or group of textual characters is not recognized by the one or more computing devices;

determining, based at least in part on relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, phrase, or group of textual characters, a plurality of entities to which the word corresponds, or

choosing, based at least in part on the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, phrase, or group of textual characters, one meaning or sense;

when the word, phrase, or group of textual characters is recognized by the one or more computing devices and the word, phrase, or group of textual characters refers to a plurality of entities, determining, with one or more computing devices, the entities to which the word, phrase, or group of textual characters corresponds;

when the word, phrase, or group of textual characters corresponds to multiple entities in a same category, ranking, with one or more computing devices, the multiple entities in the same category to which the word, phrase, or group of textual characters corresponds using a set of attributes common to entities in the category, wherein the set of attributes is different than other sets of attributes common to entities in other categories;

when the word, phrase, or group of textual characters is recognized by the one or more computing devices and the word, phrase, or group of textual characters has multiple meanings or senses, ranking, with one or more computing devices, the multiple meanings or senses of the word, phrase, or group of textual characters based on attributes of a user;

when the word, phrase, or group of textual characters is recognized by the one or more computing devices and the word, phrase, or group of textual characters has multiple meanings or senses, ranking, with one or more computing devices, a plurality of meanings or senses of the word, phrase, or group of textual characters, based on a measure of popularity of the plurality of meanings or senses;

when the word, phrase, or group of textual characters corresponds to multiple entities in a same category, or the word, phrase, or group of textual characters is recognized by the one or more computing devices and the word has multiple meanings or senses, choosing, with one or more computing devices, one meaning or sense based on the rankings;

generating, with one or more computing devices, an answer to the query based on the one meaning or sense;

providing a user interface mechanism to permit a user to select a different one of the plurality of meanings or senses, wherein the user interface mechanism visually presents a list of entities in the same category to which the word, phrase, or group of textual characters corresponds in an order determined, with one or more computing devices, according to the set of attributes common to entities in the category;

receiving, at one or more computing devices, an indication of a selected different meaning or sense selected by the user using the user interface mechanism; and

generating, with one or more computing devices, a different answer based on the selected different meaning or sense.

16. The method according to claim 15, further comprising generating a display to provide the answer to the query and to indicate the chosen meaning or sense.

17. The method according to claim 16, wherein the display indicates meanings or senses in the plurality of meanings or senses different than the chosen meaning or sense.

18. The method according to claim 17, wherein the display includes the user interface mechanism.

19. The method according to claim 17, wherein the display is a web page.

25. The method according to claim 15, further comprising:

when the word, phrase, or group of textual characters is not recognized by the one or more computing devices, generating a keyword based on an analysis of the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, phrase, or group of textual characters, wherein determining the plurality of meanings or senses is based on the keyword.

20. A tangible, non-transitory storage medium having stored thereon machine executable instructions, the machine executable instructions, when executed by one or more machines, cause the one or more machines to:

receive a query in an imprecise syntax, the query including a word, phrase, or group of textual characters;

when the word, phrase, or group of textual characters is not recognized by the one or more machines;

determine, based at least in part on relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, phrase, or group of textual characters, a plurality of entities to which the word corresponds, or

choose, based at least in part on the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, phrase, or group of textual characters, one meaning or sense;

when the word, phrase, or group of textual characters is recognized by the one or more machines and the word, phrase, or group of textual characters refers to a plurality of entities, determine the entities to which the word, phrase, or group of textual characters corresponds;

when the word, phrase, or group of textual characters corresponds to multiple entities in a same category, rank the multiple entities in the same category to which the word, phrase, or group of textual characters corresponds using a set of attributes common to entities in the category, wherein the set of attributes is different than other sets of attributes common to entities in other categories;

when the word, phrase, or group of textual characters is recognized by the one or more machines and the word, phrase, or group of textual characters has multiple meanings or senses, rank the multiple meanings or senses of the word, phrase, or group of textual characters based on attributes of a user;

when the word, phrase, or group of textual characters is recognized by the one or more machines and the word, phrase, or group of textual characters has multiple meanings or senses, rank a plurality of meanings or senses of the word, phrase, or group of textual characters, based on a measure of popularity of the plurality of meanings or senses;

when the word, phrase, or group of textual characters corresponds to multiple entities in a same category, or the word, phrase, or group of textual characters is recognized by the one or more machines and the word has multiple meanings or senses, choose one meaning or sense based on the rankings;

generate an answer to the query based on the one meaning or sense;

provide a user interface mechanism to permit a user to select a different one of the plurality of meanings or senses, wherein the user interface mechanism visually presents a list of entities in the same category to which the word, phrase, or group of textual characters corresponds in an order determined according to the set of attributes common to entities in the category;

receive an indication of a selected different meaning or sense selected by the user using the user interface mechanism; and

generate a different answer based on the selected different meaning or sense.

26. The tangible, non-transitory storage medium according to claim 20, the machine executable instructions, when executed by the one or more machines, further cause the one or more machines to:

when the word, phrase, or group of textual characters is not recognized by the one or more machines, generate a keyword based on an analysis of the relative placement of alphabetic characters, numeric characters, and non-alphanumeric characters of the word, wherein the one or more machines determines the plurality of meanings or senses based on the keyword.