Classifying text via topical analysis, for applications to speech recognition

• US 8,612,209 B2
• Filed: 09/20/2011
• Issued: 12/17/2013
• Est. Priority Date: 11/28/2002
• Status: Expired due to Term

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Helletzgruber, Matthias, Rajic, Kresimir
• Primary Examiner(s): YEN, ERIC L
• Application Number:
  US13/236,835
• Publication Number:
  US 20120010875A1
• Time in Patent Office: 819 Days
• Field of Search: 704/9, 704270-275
• US Class Current: 704/9
• CPC Class Codes: G06F 40/237 Lexical tools
• CPC Class Codes: G06F 40/268 Morphological analysis
• 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
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G16H 15/00 ICT specially adapted for medical reports, e.g. generation or transmission thereof
• CPC Class Codes: G16H 20/10 relating to drugs or medications, e.g. for ensuring correct administration to patients

Abstract

An assignment device (1) assigns word class information (WKI) to one or more words of text information (ETI). Based on word-class sequence information (WK-AI) formed from this assigned word class information (WKI), actions (A) are executed in order to notify the user of conflicts or to provide the user with background information (HI) relating to words in the text information (TT).

Claims

1. An assignment device for processing an input text comprising a first word and a second word, the assignment device comprising:

at least one computer programmed to implement a method comprising;

assigning word-class information to each of a plurality of words in the input text, including assigning first word-class information to the first word and second word-class information to the second word;

determining whether the first word-class information and the second word-class information appear within a predetermined distance of one another;

identifying, by using a mapping from combinations of word-class information to corresponding actions, an action associated with the combination of the first word-class information and the second word-class information; and

executing the identified action in response to determining that the first word-class information and the second word-class information appear within the predetermined distance of one another,

wherein the first word-class information identifies the first word as being a member of a first category of semantically related member words and the second word-class information identifies the second word as being a member of a second category of semantically related member words.

2. The assignment device of claim 1, wherein assigning word-class information to the first word comprises:

determining whether the first word is included in a word-class table that associates each of a set of words with one of a plurality of word-classes; and

when it is determined that the word is included in the word-class table, assigning the associated word-class information in the word-class table to the first word.

3. The assignment device of claim 1, wherein the identifying comprises:

determining whether the combination of the first word-class information and the second word-class information is included in a linkage table that associates a plurality of different combinations of word-class information with respective linkage information; and

when it is determined that the combination of the first word-class information and the second word-class information is included in a linkage table, determining whether the corresponding linkage information is included in an action table that associates linkage information and/or combinations of linkage information with a corresponding action.

4. The assignment device of claim 3, wherein the word-class table, the linkage table, and/or the action table can be edited by a user.

5. The assignment device of claim 1, wherein determining whether the first word-class information and the second word-class information appear within the predetermined distance of one another comprises determining whether the first word-class information and the second word-class information are within the predetermined distance of one another in word-class sequence information obtained from the input text.

6. The assignment device of claim 1, wherein determining whether the first word-class information and the second word-class information appear within the predetermined distance of one another comprises determining whether the first word and the second word appear within a threshold number of words of one another within the input text.

7. A method for processing an input text comprising a first word and a second word, the method comprising:

using at least one computer to perform the acts of;

assigning word-class information to each of a plurality of words in the input text, including assigning first word-class information to the first word and second word-class information to the second word;

determining whether the first word-class information and the second word-class information appear within a predetermined distance of one another;

identifying, by using a mapping from combinations of word-class information to corresponding actions, an action associated with the combination of the first word-class information and the second word-class information; and

executing the identified action in response to determining that the first word-class information and the second word-class information appear within the predetermined distance of one another,

wherein the first word-class information identifies the first word as being a member of a first category of semantically related member words and the second word-class information identifies the second word as being a member of a second category of semantically related member words.

8. The method of claim 7, wherein assigning word-class information to the first word comprises:

determining whether the first word is included in a word-class table that associates each of a set of words with one of a plurality of word-classes; and

when it is determined that the word is included in the word-class table, assigning the associated word-class information in the word-class table to the first word.

9. The method of claim 7, wherein the identifying comprises:

determining whether the combination of the first word-class information and the second word-class information is included in a linkage table that associates a plurality of different combinations of word-class information with respective linkage information; and

when it is determined that the combination of the first word-class information and the second word-class information is included in a linkage table, determining whether the corresponding linkage information is included in an action table that associates linkage information and/or combinations of linkage information with a corresponding action.

10. The method of claim 9, wherein the word-class table, the linkage table, and/or the action table can be edited by a user.

11. The method of claim 7, wherein determining whether the first word-class information and the second word-class information appear within the predetermined distance of one another comprises determining whether the first word-class information and the second word-class information are within the predetermined distance of one another in word-class sequence information obtained from the input text.

12. The method of claim 7, wherein determining whether the first word-class information and the second word-class information appear within the predetermined distance of one another comprises determining whether the first word and the second word appear within a threshold number of words of one another within the input text.

13. At least one non-transitory computer readable medium which comprises sections of software code, executable by at least one computer, that when executed by the at least one computer cause the at least one computer to perform a method for processing an input text comprising a first word and a second word, the method comprising:

assigning word-class information to each of a plurality of words in the input text, including assigning first word-class information to the first word and second word-class information to the second word;

determining whether the first word-class information and the second word-class information appear within a predetermined distance of one another;

identifying, by using a mapping from combinations of word-class information to corresponding actions, an action associated with the combination of the first word-class information and the second word-class information; and

executing the identified action in response to determining that the first word-class information and the second word-class information appear within the predetermined distance of one another,

wherein the first word-class information identifies the first word as being a member of a first category of semantically related member words and the second word-class information identifies the second word as being a member of a second category of semantically related member words.

14. The at least one non-transitory computer readable medium of claim 13, wherein assigning word-class information to the first word comprises:

determining whether the first word is included in a word-class table that associates each of a set of words with one of a plurality of word-classes; and

when it is determined that the word is included in the word-class table, assigning the associated word-class information in the word-class table to the first word.

15. The at least one non-transitory computer readable medium of claim 13, wherein the identifying comprises:

determining whether the combination of the first word-class information and the second word-class information is included in a linkage table that associates a plurality of different combinations of word-class information with respective linkage information; and

when it is determined that the combination of the first word-class information and the second word-class information is included in a linkage table, determining whether the corresponding linkage information is included in an action table that associates linkage information and/or combinations of linkage information with a corresponding action.

16. The at least one non-transitory computer readable medium of claim 15, wherein the word-class table, the linkage table, and/or the action table can be edited by a user.

17. The at least one non-transitory computer readable medium of claim 13, wherein determining whether the first word-class information and the second word-class information appear within the predetermined distance of one another comprises determining whether the first word-class information and the second word-class information are within the predetermined distance of one another in word-class sequence information obtained from the input text.

18. The at least one non-transitory computer readable medium of claim 13, wherein determining whether the first word-class information and the second word-class information appear within the predetermined distance of one another comprises determining whether the first word and the second word appear within a threshold number of words of one another within the input text.