Intention detection in domain-specific information

• US 9,355,089 B1
• Filed: 12/08/2014
• Issued: 05/31/2016
• Est. Priority Date: 12/08/2014
• Status: Expired due to Fees

• Current Assignee: Baylor College of Medicine (Baylor College), International Business Machines Corporation, Board of Regents of the University of Texas System (University of Texas System)
• Original Assignee: International Business Machines Corporation
• Inventors: Bachman, Benjamin J., Donehower, Lawrence A., Regenbogen, Sam J., Pickering, Curtis R., Nagarajan, Meenakshi, Spangler, William Scott, Lichtarge, Olivier, Wilkins, Angela D.
• Primary Examiner(s): Augustin, Marcellus
• Application Number:
  US14/564,068
• Publication Number:
  US 20160162465A1
• Time in Patent Office: 540 Days
• Field of Search: 704/4, 704/7, 704/222, 704/236, 704/237, 704/243, 704/245, 704/251
• US Class Current: 1/1
• CPC Class Codes: G06F 16/313 Selection or weighting of terms for indexing
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/253 Grammatical analysis; Style critique
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates
• CPC Class Codes: G06F 40/289 Phrasal analysis, e.g. finite state techniques or chunking

Abstract

A new information in a language and relating to a subject matter domain is parsed into a constituent set of complete grammatical constructs. In a subset of the complete grammatical constructs, a set of linguistic styles of the language is identified according to a subset of a set of word-style associations related to the language and independent of the subject matter domain. A first weight is assigned to a first linguistic style and a second weight to a second linguistic style from the set of linguistic styles. A first intention information is mapped to the first style using a first style-intention rule, and a second intention information to the second style using a second style-intention rule. A complete grammatical construct in the subset is tagged with the first intention information responsive to a weight associated with the first intention information exceeding an intention selection threshold.

Claims

1. A method for intention detection in information, the method comprising:

parsing, using a processor and a memory, a new information into a constituent set of complete grammatical constructs, the new information being in a language and relating to a subject matter domain;

identifying, in a subset of the complete grammatical constructs, a set of linguistic styles of the language according to a subset of a set of word-style associations related to the language and independent of the subject matter domain;

assigning a first weight to a first linguistic style and a second weight to a second linguistic style, the set of linguistic styles including the first linguistic style and the second linguistic style;

constructing, using second set of linguistic styles of the language, and using a taxonomy of intentions present in a training information relating to the subject matter domain, a set of style-intention rules, the set of style-intention rules including the style-intention rule, the style-intention rule relating a selected intention from the taxonomy of intentions with a selected linguistic style in the set of linguistic styles to indicate that the selected intention is present where the selected linguistic style appears in any information related to the subject matter domain;

mapping a first intention information to the first style using a first style-intention rule, and a second intention information to the second style using a second style-intention rule; and

tagging a complete grammatical construct in the subset with the first intention information responsive to a weight associated with the first intention information exceeding an intention selection threshold.

2. The method of claim 1, further comprising:

determining, using an encoded structure, that the subset of complete grammatical constructs is related to a section of the encoded structure, wherein the new information conforms to the encoded structure;

determining, using the encoded structure, a second set of linguistic styles corresponding to the section of the encoded structure;

concluding that the first linguistic style and the second linguistic style are also members of the second set of linguistic styles, wherein the assigning the first second and third weights is responsive to the concluding.

3. The method of claim 2, further comprising:

evaluating, using the encoded structure, that the first linguistic style is more likely to be present in the section than the second linguistic style, wherein the first weight is greater than the second weight responsive to the evaluating.

4. The method of claim 1, further comprising:

assigning a third weight to a third linguistic style, the set of linguistic styles including the third linguistic style; and

omitting the third linguistic style from the mapping responsive to the third weight being below a style selection threshold.

5. The method of claim 1, further comprising:

constructing, using a training information, a structure, wherein the training information conforms to the structure;

encoding the structure, to form the encoded structure, by associating a third linguistic style that is present in a portion of the training information with a section of the structure; and

assigning a third weight to the third linguistic style in the section.

6. The method of claim 1, further comprising:

constructing, using a lexicon of the language, the set of word-style associations, wherein a word in a word-style association is used in the language to convey the corresponding style in the word-style association.

7. The method of claim 1, further comprising:

adjusting, according to a third style-intention rule, the first weight to a third weight, wherein the first intention information is associated with the third weight.

8. The method of claim 1, further comprising:

omitting tagging the complete grammatical construct in the subset with the second intention information responsive to a weight associated with the second intention information not exceeding the intention selection threshold.

9. The method of claim 1, wherein the language is English, and a complete grammatical construct in the set of complete grammatical constructs is a sentence according to English grammar.

10. A computer usable program product comprising a computer readable storage device including computer usable code for intention detection in information,

the computer usable code comprising:

computer usable code for parsing a new information into a constituent set of complete grammatical constructs, the new information being in a language and relating to a subject matter domain;

computer usable code for identifying, in a subset of the complete grammatical constructs, a set of linguistic styles of the language according to a subset of a set of word-style associations related to the language and independent of the subject matter domain;

computer usable code for assigning a first weight to a first linguistic style and a second weight to a second linguistic style, the set of linguistic styles including the first linguistic style and the second linguistic style;

computer usable code for constructing, using second set of linguistic styles of the language, and using a taxonomy of intentions present in a training information relating to the subject matter domain, a set of style-intention rules, the set of style-intention rules including the style-intention rule, the style-intention rule relating a selected intention from the taxonomy of intentions with a selected linguistic style in the set of linguistic styles to indicate that the selected intention is present where the selected linguistic style appears in any information related to the subject matter domain;

computer usable code for mapping a first intention information to the first style using a first style-intention rule, and a second intention information to the second style using a second style-intention rule;

and computer usable code for tagging a complete grammatical construct in the subset with the first intention information responsive to a weight associated with the first intention information exceeding an intention selection threshold.

11. The computer usable program product of claim 10, further comprising:

computer usable code for determining, using an encoded structure, that the subset of complete grammatical constructs is related to a section of the encoded structure, wherein the new information conforms to the encoded structure;

computer usable code for determining, using the encoded structure, a second set of linguistic styles corresponding to the section of the encoded structure;

computer usable code for concluding that the first linguistic style and the second linguistic style are also members of the second set of linguistic styles, wherein the assigning the first second and third weights is responsive to the concluding.

12. The computer usable program product of claim 11, further comprising:

computer usable code for evaluating, using the encoded structure, that the first linguistic style is more likely to be present in the section than the second linguistic style, wherein the first weight is greater than the second weight responsive to the evaluating.

13. The computer usable program product of claim 10, further comprising:

computer usable code for assigning a third weight to a third linguistic style, the set of linguistic styles including the third linguistic style; and

computer usable code for omitting the third linguistic style from the mapping responsive to the third weight being below a style selection threshold.

14. The computer usable program product of claim 10, further comprising:

computer usable code for constructing, using second set of linguistic styles of the language, and using a taxonomy of intentions present in a training information relating to the subject matter domain, a set of style-intention rules, the set of style-intention rules including the style-intention rule, the style-intention rule relating a selected intention from the taxonomy of intentions with a selected linguistic style in the set of linguistic styles to indicate that the selected intention is present where the selected linguistic style appears in any information related to the subject matter domain.

15. The computer usable program product of claim 10, further comprising:

computer usable code for constructing, using a training information, a structure, wherein the training information conforms to the structure;

computer usable code for encoding the structure, to form the encoded structure, by associating a third linguistic style that is present in a portion of the training information with a section of the structure; and

computer usable code for assigning a third weight to the third linguistic style in the section.

16. The computer usable program product of claim 10, further comprising:

computer usable code for constructing, using a lexicon of the language, the set of word-style associations, wherein a word in a word-style association is used in the language to convey the corresponding style in the word-style association.

17. The computer usable program product of claim 10, wherein the computer usable code is stored in a computer readable storage device in a data processing system, and wherein the computer usable code is transferred over a network from a remote data processing system.

18. The computer usable program product of claim 10, wherein the computer usable code is stored in a computer readable storage device in a server data processing system, and wherein the computer usable code is downloaded over a network to a remote data processing system for use in a computer readable storage device associated with the remote data processing system.

19. A data processing system for intention detection in information,

the data processing system comprising:

a storage device, wherein the storage device stores computer usable program code;

and a processor, wherein the processor executes the computer usable program code, and wherein the computer usable program code comprises;

computer usable code for parsing a new information into a constituent set of complete grammatical constructs, the new information being in a language and relating to a subject matter domain;

computer usable code for identifying, in a subset of the complete grammatical constructs, a set of linguistic styles of the language according to a subset of a set of word-style associations related to the language and independent of the subject matter domain;

computer usable code for assigning a first weight to a first linguistic style and a second weight to a second linguistic style, the set of linguistic styles including the first linguistic style and the second linguistic style;

computer usable code for constructing, using second set of linguistic styles of the language, and using a taxonomy of intentions present in a training information relating to the subject matter domain, a set of style-intention rules, the set of style-intention rules including the style-intention rule, the style-intention rule relating a selected intention from the taxonomy of intentions with a selected linguistic style in the set of linguistic styles to indicate that the selected intention is present where the selected linguistic style appears in any information related to the subject matter domain;

computer usable code for mapping a first intention information to the first style using a first style-intention rule, and a second intention information to the second style using a second style-intention rule;

and computer usable code for tagging a complete grammatical construct in the subset with the first intention information responsive to a weight associated with the first intention information exceeding an intention selection threshold.