Context-aware question answering system

• US 8,935,277 B2
• Filed: 03/30/2012
• Issued: 01/13/2015
• Est. Priority Date: 03/30/2012
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: SAP SE
• Inventors: Kuchmann-Beauger, Nicolas, Aufaure, Marie-Aude, Thollot, Raphael
• Primary Examiner(s): Nguyen, Loan T
• Application Number:
  US13/434,899
• Publication Number:
  US 20130262501A1
• Time in Patent Office: 1,019 Days
• US Class Current: 707/769
• CPC Class Codes: G06F 16/24535 of sub-queries or views
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence

Abstract

A question is received to be answered by a question answering (QA) system. The question may be a business intelligence question that is expressed in a natural language. The question is parsed. The parsed question is matched to a pattern from a number of patterns. A technical query associated with the matched pattern is processed to retrieve data relevant to the question from a number of data sources. The QA system generates an answer to the question based on retrieved data. In one aspect, the QA system generates answers based contextual information.

Claims

1. A computer implemented method to generate an answer to a business intelligence (BI) request for information, the method comprising:

receiving the BI request for information expressed in natural language, the BI request is sent by an agent;

upon receiving the BI request, loading into a memory a situation graph of the agent, the situation graph representing information contextual to the agent;

parsing the BI request;

based on the parsed BI request, generating a graph representing syntactic structure of the received BI request;

enriching the graph based on the parsed BI request with;

at least one semantic annotation representing at least one business object identified in the BI request, the business object included in a data model that at least in part forms a semantic layer of a plurality of data sources, the data model determines a structure of at least one data source from the plurality of data sources, and

at least one semantic annotation representing contextual information derived from the situation graph of the agent;

matching the parsed BI request to a pattern from a plurality of patterns of features included in the BI request, wherein features of the pattern include reference to the at least one business object identified in the BI request;

processing by the computer a technical query associated with the pattern from the plurality of patterns to retrieve data relevant to the BI request at least from the at least one data source

generating the answer to the BI request based on the retrieved data relevant to the BI request and based at least in part on the situation graph of the agent, wherein the answer generated by triggering at least one operator of a situational recommender system; and

recommending the generated answer.

2. The method of claim 1, wherein recommending the answer further comprises:

creating a visualization of the answer based on the retrieved data.

3. The method of claim 1, further comprising:

generating an answer to a follow-up request for information, wherein the follow-up request for information refers to the BI request for information.

4. The method of claim 1, further comprising:

ranking one or more generated answers, and

adding the ranked one or more answers to the situation graph of the agent, wherein the one or more answers correspond to the BI request for information.

5. The method of claim 1, wherein recommending the answer further comprises:

resolving ambiguities in the BI request for information based on the situation graph of the agent.

6. A question-answering computer system to generate an answer to a business intelligence (BI) request for information, the system including:

at least one processor and memory to execute program code related to;

receive the BI request for information expressed in natural language, the BI request is sent by an agent;

upon receiving the BI request, loading into a memory a situation graph of the agent, the situation graph representing information contextual to the agent;

parse the BI request;

based on the parsed BI request, generate a graph representing syntactic structure of the received BI request;

enrich the graph based on the parsed BI request;

at least one semantic annotation representing at least one business object identified in the BI request, the business object included in a data model that at least in part forms semantic layer of a plurality of data sources, the data model determines a structure of at least one data source from the plurality of data sources,

at least one semantic annotation representing contextual information derived from the situation graph of the agent;

match the parsed BI request to a pattern from a plurality of patterns of features included in the BI request, wherein features of the pattern include reference to the at least one business object identified in the BI request;

process by the computer a technical query associated with the pattern from the plurality of patterns to retrieve data relevant to the BI request at least from the at least one data source;

generate the answer to the BI request for information based on the retrieved data relevant to the BI request and based at least in part on the situation graph of the agent, wherein the answer generated by triggering at least one operator of a situational recommender system; and

recommending the generated answer.

7. The system of claim 6, wherein recommending the answer further comprises:

rendering a visualization of the answer based on the retrieved data.

8. The system of claim 6, further comprising:

generating an answer to a follow-up request for information, wherein the follow-up request for information refers to the BI request for information.

9. The system of claim 6, further comprising:

ranking one or more generated answers, and

adding the ranked one or more answers to the situation graph of the agent, wherein the one or more answers correspond to the BI request for information.

10. The system of claim 6, wherein recommending the answer further comprises:

resolving ambiguities in the BI request for information based on the situation graph of the agent.

11. A non-transitory computer readable medium storing instructions thereon, which when executed by a processor cause a computer system to:

receive a business intelligence (BI) request for information expressed in natural language, the BI request is sent by an agent;

upon receiving the BI request, loading into a memory a situation graph of the agent, the situation graph representing information contextual to the agent;

parse the BI request;

based on the parsed BI request, generate a graph representing syntactic structure of the received BI request;

enrich the graph based on the parsed BI request with;

at least one semantic annotation representing at least one business object identified in the BI request, the business object included in a data model that at least in part forms a semantic layer of a plurality of data sources, the data model determines a structure of at least one data source from the plurality of data sources, and

at least one semantic annotation representing contextual information derived from the situation graph of the agent;

match the parsed BI request for information to a pattern from a plurality of patterns of features included in the BI request, wherein features of the pattern include reference to the at least one business object identified in the BI request;

process by the computer a technical query associated with the pattern from the plurality of patterns to retrieve data relevant to the BI request at least from the at least one data source;

generate the answer to the BI request based on the retrieved data relevant to the BI request and based at least in part on the situation graph of the agent, wherein the answer generated by triggering at least one operator of a situational recommender system; and

recommending the generated answer.

12. The non-transitory computer readable medium of claim 11, further comprising:

generating an answer to a follow-up request for information, wherein the follow-up request for information refers to the BI request for information.

13. The non-transitory computer readable medium of claim 11, further comprising:

ranking one or more generated answers, and

adding the ranked one or more answers to the situation graph of the agent, wherein the one or more answers correspond to the BI request for information.

14. The non-transitory computer readable medium of claim 11, wherein recommending the answer further comprises:

resolving ambiguities in the BI request for information based on the situation graph of the agent.