Classifying text via topical analysis, for applications to speech recognition

• US 8,032,358 B2
• Filed: 10/31/2003
• Issued: 10/04/2011
• Est. Priority Date: 11/28/2002
• Status: Expired due to Fees

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Nuance Communications Austria Gmbh (Microsoft Corporation)
• Inventors: Rajic, Kresimir, Helletzgruber, Matthias
• Primary Examiner(s): Yen; Eric
• Application Number:
  US10/536,239
• Publication Number:
  US 20060074668A1
• Time in Patent Office: 2,895 Days
• Field of Search: 704/9, 704/257, 704/231
• 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 (TI).

Claims

1. An assignment device for obtaining meaning from an input text to identify an appropriate action to be taken based on the meaning, the assignment device comprising:

at least one computer programmed to implement;

an assignment component configured to access a word-class table associating each of a set of words with one of a plurality of word-classes to which the respective word is a member, each of the plurality of word-classes characterizing a categorical commonality of all member words related to the definitional meaning of the member words, the assignment component determining whether each word in the input text is included in the word-class table and assigning word-class information to each word in the input text that is included in the word-class table that indicates to which word-class the corresponding word is a member, and providing word-class sequence information containing the assigned word-class information in the sequence in which the associated words appear in the input text;

a linkage component configured to access a linkage table that associates a plurality of different combinations of word-class information with respective linkage information, the linkage component configured to detect if any of the plurality of combinations of word-class information are present in the word-class sequence information, and provide the associated linkage information for each detected combination; and

an action component to receive the provided linkage information and configured to access an action table that associates linkage information and/or combinations of linkage information with a corresponding action, the action component configured to access the action table to detect if any action is associated with the linkage information received from the linkage component, and to activate each action detected as being associated with the linkage information.

2. An assignment device as claimed in claim 1, wherein the linkage component is designed to only provide the linkage information associated with a detected word-class combination if the corresponding words are less than a maximum distance from one another in the input text distance.

3. An assignment device as claimed in claim 1, wherein the word-class table includes a medicament class and the linkage table includes at least one combination of the medicament class and another of the plurality of word-classes, and wherein the action table includes at least one action including providing an alert when the input text includes a member of the medicament class and a member of the another of the plurality of word-classes.

4. An assignment device as claimed in claim 1, wherein the assignment component is part of a voice recognition device and is designed to assign word-class information to one or more words of the input text recognized by a voice recognition component.

5. An assignment device as claimed in claim 1, wherein the action component is designed to activate an acoustic and/or visual notification as at least one of the actions.

6. An assignment device as claimed in claim 1, wherein the word-class table, the linkage table and the action table can be manually adjusted.

7. An assignment device as claimed in claim 1, wherein at least one action in the action table instructs the action component to establish and to deliver background information that can be displayed with associated words in the input text.

8. An assignment method for processing an input text, the method comprising acts of:

accessing, using at least one computer, a word-class table associating each of a set of words with one of a plurality of word-classes to which the respective word is a member, each of the plurality of word-classes characterizing a categorical commonality of all member words related to the definitional meaning of the member words;

determining, using at least one computer, whether each word in the input text is included in the word-class table and assigning word-class information to each word in the input text that is included in the word-class table that indicates to which word-class the corresponding word is a member;

delivering, using at least one computer, word-class sequence information containing the assigned word-class information in the sequence in which the associated words appear in the input text;

accessing, using at least one computer, a linkage table that associates a plurality different combinations of word-class information of interest with respective linkage information;

detecting, using at least one computer, if any of the plurality of combinations of word-class information are present in the word-class information;

providing, by at least one computer, the associated linkage information for each detected combination;

receiving, by at least one computer, the provided linkage information;

accessing, using at least one computer, an action table that associates linkage information and/or combinations of linkage information with a corresponding action;

detecting, using at least one computer, if any action is associated with the linkage information; and

activating, using at least one computer, each action detected as being associated with the linkage information.

9. An assignment method as claimed in claim 8, wherein the linkage information associated with a detected word-class combination is delivered only if the corresponding words are less than a maximum distance from one another in the input text.

10. An assignment method as claimed in claim 8, wherein the input text is obtained using a voice recognition method.

11. An assignment method as claimed in claim 8, wherein an acoustic and/or visual notification is delivered as at least one of the actions in the action table.

12. An assignment method as claimed in claim 8, wherein at least one action includes delivering background information read from a background dictionary relating to the words that triggered the at least one action.

13. A non-transitory computer readable medium which comprises sections of software code, executable by computer, the sections of software code, when executed by the computer performing a method for processing an input text, the method comprising:

accessing a word-class table associating each of a set of words with one of a plurality of word-classes to which the respective word is a member, each of the plurality of word-classes characterizing a categorical commonality of all member words related to the definitional meaning of the member words;

determining whether each word in the input text is included in the word-class table and assigning word-class information to each word in the input text that is included in the word-class table that indicates to which word-class the corresponding word is a member;

delivering word-class sequence information containing the assigned word-class information in the sequence in which the associated words appear in the input text;

accessing a linkage table that associates a plurality different combinations of word-class information of interest with respective linkage information;

detecting if any of the plurality of combinations of word-class information are present in the word-class information;

providing the associated linkage information for each detected combination;

receiving the provided linkage information;

accessing an action table that associates linkage information and/or combinations of linkage information with a corresponding action; and

detecting if any action is associated with the linkage information; and

activating each action detected as being associated with the linkage information.