Question answering from structured and unstructured data sources

Patent metadata

Field	Value
Publication number	US-2016179934-A1
Application number	US-201414575404-A
Country	US
Kind code	A1
Filing date	Dec 18, 2014
Priority date	Dec 18, 2014
Publication date	Jun 23, 2016
Grant date	—

Abstract

Techniques for question answering involve receiving, from a user, a text input expressing a question in natural language. The text input may be analyzed, including identifying in the question at least one first portion answerable from at least one structured data source, and at least one second portion answerable from at least one unstructured data source. At least one first query configured for the structured data source(s) may be constructed from the at least one first portion of the question and applied to the structured data source(s) to retrieve first answer information for the at least one first portion of the question. At least one second query configured for the unstructured data source(s) may be constructed from the at least one second portion of the question and applied to the unstructured data source(s) to retrieve second answer information for the at least one second portion of the question.

First claim

What is claimed is:

1 . A method comprising:
receiving, from a user, a text input expressing a question in natural language; analyzing the text input, using an analysis component implemented via at least one processor, the analyzing comprising identifying in the question at least one first portion answerable from at least one structured data source, and at least one second portion answerable from at least one unstructured data source; constructing, from the at least one first portion of the question, at least one first query configured for the at least one structured data source, and applying the at least one first query to the at least one structured data source to retrieve first answer information for the at least one first portion of the question; and constructing, from the at least one second portion of the question, at least one second query configured for the at least one unstructured data source, and applying the at least one second query to the at least one unstructured data source to retrieve second answer information for the at least one second portion of the question.

2 . The method of claim 1 , further comprising:
merging the first answer information from the at least one structured data source and the second answer information from the at least one unstructured data source to form an answer to the question; and presenting the answer to the user.

3 . The method of claim 1 , wherein constructing the at least one second query comprises using the first answer information to constrain the at least one second query.

4 . The method of claim 1 , wherein the at least one structured data source comprises at least one database.

5 . The method of claim 1 , wherein the at least one unstructured data source comprises at least one set of documents comprising natural language text.

6 . The method of claim 5 , further comprising:
analyzing at least one document in the at least one set of documents comprising natural language text to identify at least one section in the at least one document as being relevant to at least one classification category appearing in the at least one structured data source, and generating at least one annotation identifying the at least one section as being relevant to the at least one classification category; wherein applying the at least one second query comprises applying the at least one second query at least in part to the at least one annotation.

7 . The method of claim 5 , further comprising:
identifying, in the at least one set of documents comprising natural language text, at least one portion of natural language text as providing evidence that supports the second answer information; and presenting the at least one portion of natural language text to the user in association with an answer to the question.

8 . At least one computer-readable storage medium storing computer-executable instructions that, when executed, perform a method comprising:
receiving, from a user, a text input expressing a question in natural language; analyzing the text input, the analyzing comprising identifying in the question at least one first portion answerable from at least one structured data source, and at least one second portion answerable from at least one unstructured data source; constructing, from the at least one first portion of the question, at least one first query configured for the at least one structured data source, and applying the at least one first query to the at least one structured data source to retrieve first answer information for the at least one first portion of the question; and constructing, from the at least one second portion of the question, at least one second query configured for the at least one unstructured data source, and applying the at least one second query to the at least one unstructured data source to retrieve second answer information for the at least one second portion of the question.

9 . The at least one computer-readable storage medium of claim 8 , wherein the method further comprises:
merging the first answer information from the at least one structured data source and the second answer information from the at least one unstructured data source to form an answer to the question; and presenting the answer to the user.

10 . The at least one computer-readable storage medium of claim 8 , wherein constructing the at least one second query comprises using the first answer information to constrain the at least one second query.

11 . The at least one computer-readable storage medium of claim 8 , wherein the at least one structured data source comprises at least one database.

12 . The at least one computer-readable storage medium of claim 8 , wherein the at least one unstructured data source comprises at least one set of documents comprising natural language text.

13 . The at least one computer-readable storage medium of claim 12 , wherein the method further comprises:
analyzing at least one document in the at least one set of documents comprising natural language text to identify at least one section in the at least one document as being relevant to at least one classification category appearing in the at least one structured data source, and generating at least one annotation identifying the at least one section as being relevant to the at least one classification category; wherein applying the at least one second query comprises applying the at least one second query at least in part to the at least one annotation.

14 . The at least one computer-readable storage medium of claim 12 , wherein the method further comprises:
identifying, in the at least one set of documents comprising natural language text, at least one portion of natural language text as providing evidence that supports the second answer information; and presenting the at least one portion of natural language text to the user in association with an answer to the question.

15 . Apparatus comprising:
at least one processor; and at least one storage medium storing processor-executable instructions that, when executed by the at least one processor, perform a method comprising:
receiving, from a user, a text input expressing a question in natural language;
analyzing the text input, the analyzing comprising identifying in the question at least one first portion answerable from at least one structured data source, and at least one second portion answerable from at least one unstructured data source;
constructing, from the at least one first portion of the question, at least one first query configured for the at least one structured data source, and applying the at least one first query to the at least one structured data source to retrieve first answer information for the at least one first portion of the question; and
constructing, from the at least one second portion of the question, at least one second query configured for the at least one unstructured data source, and applying the at least one second query to the at least one unstructured data source to retrieve second answer information for the at least one second portion of the question.

16 . The apparatus of claim 15 , wherein the method further comprises:
merging the first answer information from the at least one structured data source and the second answer information from the at least one unstructured data source to form an answer to the question; and presenting the answer to the user.

17 . The apparatus of claim 15 , wherein constructing the at least one second query comprises using the first answer information to constrain the at least one

Assignees

• Nuance Communications Inc

Inventors

• Stubley Peter
• Boston Marisa Ferrara
• Crouch Richard Stamford
• Ozcan Ali Erdem

Classifications

• G06F17/30696

  Physics · mapped topic

• G06F17/30401

  Physics · mapped topic

• G06F17/30654 Primary

  Physics · mapped topic

• G06F17/30684

  Physics · mapped topic

• G06N5/00

  Computing arrangements using knowledge-based models · CPC title