System & method for processing sentence based queries

• US 20050086049A1
• Filed: 12/03/2004
• Published: 04/21/2005
• Est. Priority Date: 11/12/1999
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Phoenix Solutions Incorporated (CDC Corporation)
• Inventors: Bennett, Ian M.
• Granted Patent: US 7,657,424 B2
• Time in Patent Office: Days
• US Class Current: 704/4
• CPC Class Codes: G06F 16/243 Natural language query formulation
• CPC Class Codes: G06F 16/24522 Translation of natural language queries to structured queries
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/216 using statistical methods
• CPC Class Codes: G06F 40/237 Lexical tools
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/42 Data-driven translation
• CPC Class Codes: G06F 40/44 Statistical methods, e.g. probability models
• CPC Class Codes: G09B 5/04 with audible presentation of the material to be studied sound recording or reproducing G11B
• CPC Class Codes: G09B 7/00 Electrically-operated teaching apparatus or devices working with questions and answers mechanically operated G09B3/00; computing arrangements G06F
• CPC Class Codes: G10L 15/142 Hidden Markov Models [HMMs]
• CPC Class Codes: G10L 15/18 using natural language modelling
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/285 Memory allocation or algorithm optimisation to reduce hardware requirements
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: H04M 2250/74 with voice recognition means voice activated dialling H04M1/271; voice recognition algorithms G10L15/00
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

Sentence based queries from a user are analyzed to determine appropriate answers from an electronic database. Text in the queries is dissected to identify different phrases; the phrases in turn are matched against candidate phrases to determine best matching answers responsive to the user'"'"'s question. The system and methods are useful for Internet based search engines, as well as distributed speech recognition systems such as a client-server system. The latter are typically implemented on an intranet or over the Internet based on user queries at his/her computer, a PDA, or a workstation using a speech input interface.

Claims

1. An electronic query response system comprising:

a first software routine configured to form a text string from text words associated with a user query;

a first database storing responsive answer data items which can be presented in response to said user query;

a second software routine configured to generate a database query to said first database based on said text string;

wherein said second software routine causes said first database to retrieve and identify a candidate set of responsive answer data items for said database query;

a third software routine configured to analyze said text string and text within each of said candidate set of responsive answer data items to compare a first list of word phrases within said text string with corresponding second lists of word phrases within each of each of said candidate set of responsive answer data items;

wherein said third software routine can determine which of said candidate set of responsive answer data items best corresponds to said user query.

2. The system of claim 1, wherein said second software routine generates a first level database query using only said text string and further customizes said first level database query using a natural language evaluation of said text string.

3. The system of claim 2, wherein said natural language evaluation includes statistical processing to identify semantic variants of words in said text string.

4. The system of claim 2, wherein said natural language evaluation operates during a time when said second software routine generates said first level database query.

5. The system of claim 2 wherein said natural language evaluation further includes parts of speech (POS) analysis of said user query.

6. The system of claim 5 wherein said natural language evaluation further includes linguistically processing said candidate set of responsive answer data items to determine a final match for said user query.

7. The system of claim 1, wherein said user query is presented within a search engine operating on an Internet based Web page.

8. The system of claim 1, wherein a set of context specific dictionaries are used for recognizing said user query.

9. The system of claim 2, wherein said natural language evaluation includes an operation for determining each noun-phrase presented within said list of word phrases contained in said text words.

10. The system of claim 2, wherein said natural language evaluation is performed by a natural language engine adapted for a particular native language used by a user providing said user query.

11. The system of claim 1, wherein a ranked array of individual word phrases is compiled from said first list of word phrases and/or said second first list of word phrases and used for determining which of said candidate set of responsive answer data items best corresponds to said user query.

12. The system of claim 1, wherein said user query is matched to one or more of said candidate set of responsive answer data items by comparing a number of individual word phrases which are in common between said first list of word phrases and corresponding ones of said second lists of word phrases.

13. The system of claim 12, wherein said word phrases are noun phrases generated by a natural language engine.

14. The system of claim 13, wherein said noun phrases are determined by grouping tagged token words presented in said text words.

15. The system of claim 1, wherein said candidate set of responsive answer data items are stored in paired form with predefined user queries.

16. The system of claim 1, wherein said user query is derived from a recognized speech utterance.

17. The system of claim 1, wherein said database query is a full text query containing SQL search predicates.

18. A query processing system adapted for assisting recognition of a user query, the system comprising:

a text processing routine adapted to receive text from a search engine representing the user query and to generate a first plurality of individual word phrases derived from said text;

a database query routine adapted to process said text and formulate a search request to a database based on a list of words contained in said text;

wherein said database query routine retrieves a candidate set of potential matches for the user query; and

said text processing routine being further configured for;

(i) performing analyses of text contained in said candidate set of potential matches to determine a corresponding set of a second plurality of individual word phrases therein; and

(ii) comparing said first first plurality of individual word phrases with each of said corresponding set of a second plurality of individual word phrases to identify a match for said user query.

19. The system of claim 18, wherein said text processing includes an operation to determine noun-phrases.

20. The system of claim 19, wherein said match is based on comparing a set of noun-phrases found in each of said candidate set of potential matches and a separate set of noun-phrases in said user query.

21. The system of claim 19, wherein the text processing routine includes a natural language routine which performs statistical processing to identify semantic variants of words in said text.

22. A method of recognizing a query comprising the steps of:

(a) receiving a text string representing the query; and

(b) processing said text string to generate first word phrases and associated search predicates corresponding to the query;

(c) generating a database query to identify one or more potential matches for the query, said database query being based on said text string and said search predicates;

(d) determining a match for the query by comparing second word phrases within said one or more potential matches against said first word phrases.

23. The method of claim 22, further including a step:

(e) retrieving a matching response for said final match, which matching response is provided in audible form.

24. The method of claim 22, wherein steps (b) and (c) overlap in time.

25. The method of claim 24, wherein step (c) includes two sub-steps, including a step (c)′

wherein a preliminary query is generated based on said text string, and a step (c)″

wherein a final query is generated based on said preliminary query and said search predicates.

26. The method of claim 22, wherein said final match is determined by comparing noun-phrases.

27. The method of claim 22, wherein step (a) occurs across a distributed computing platform, including a client device and a server device, and said steps (a) through (d) are divided according to resources available at said client device and said server device respectively.

28. The method of claim 22, wherein steps (a) to (d) occur simultaneously across multiple servers and/or associated servers configured in a network farm in response to a speech utterance from a single client device.