Question Answering Entailment Analysis Structured Unstructured Data

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional question answering systems are limited in their ability to effectively address questions that require information from both structured and unstructured data sources, as they often rely solely on one type of data source, which can lead to incomplete or inaccurate answers.

Innovation Solution

A method that analyzes user input questions to identify portions answerable from structured and unstructured data sources separately, constructs appropriate queries for each, and merges the retrieved information to provide comprehensive answers, utilizing entailment analysis to determine supporting evidence from unstructured data sources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a QA system relies solely on structured data sources, then query processing is efficient and straightforward, but the system cannot answer questions requiring unstructured information

Engineering Contradiction:

Improveability to answer questions from multiple data typesVSAvoidsystem architecture complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system divides the question analysis process into distinct segments: identifying structured data portions and unstructured data portions separately. Each portion is then processed through appropriate query construction methods tailored to its data type, allowing the system to handle multiple data sources without overwhelming complexity

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The question analysis component serves multiple functions: it analyzes the input question, identifies portions answerable from structured data, identifies portions answerable from unstructured data, and coordinates both query types. This multi-functional approach consolidates what would otherwise require separate systems

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Loss of information

If a QA system uses only unstructured data sources, then comprehensive information retrieval is possible, but answer precision and reliability decrease

Engineering Contradiction:

Improvecompleteness of answer informationVSAvoidanswer accuracy

Core Design Contradiction:

Loss of informationVSMeasurement precision

Solution Approach 1:

The system merges results from structured data queries and unstructured data queries into a unified answer. Structured data provides precise, reliable information while unstructured data supplements with comprehensive context, and the merging process integrates both strengths to produce complete and accurate answers

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The system introduces an intermediary analysis layer that evaluates both structured and unstructured data results, determining which portions answer the question and how they complement each other. This intermediary process ensures that precision from structured data and completeness from unstructured data are both preserved in the final answer

Inventive Principle:

Principle #24Intermediary (Mediator)

3Productivity

If a QA system performs keyword searches on web pages, then document retrieval is simple and fast, but the user must manually examine each document to find the answer

Engineering Contradiction:

Improveanswer retrieval efficiencyVSAvoiduser effort to locate answer

Core Design Contradiction:

ProductivityVSEase of operation

Solution Approach 1:

The system performs self-service by automatically analyzing the input question, identifying answerable portions, constructing appropriate queries, retrieving relevant information, and presenting the answer directly to the user. This eliminates the need for users to manually examine multiple documents, as the system independently completes the entire answer-finding process

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS10783159B2Question answering with entailment analysis

Publication Date: 2020.09.22 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Techniques for question answering involve receiving, from a user, a text input expressing a question in natural language. In response to the question, a text output expressing an answer to the question may be generated. A plurality of documents comprising natural language text may be analyzed, involving mapping the question to one or more hypotheses, analyzing at least one passage of text in at least one of the documents to determine whether the passage entails at least one of the hypotheses, and in response to determining that the passage entails at least one of the hypotheses, identifying the passage as providing supporting evidence for the answer to the question. The answer and the at least one passage identified as providing supporting evidence for the answer may be presented to the user in response to the text input.