Natural Language Query Processing via Knowledge Model Assertions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional information retrieval systems face challenges in achieving high precision and recall when handling natural language queries, often resulting in ambiguous results due to the trade-off between the two metrics, especially in large databases with many irrelevant documents.
Innovation Solution
An information retrieval system that uses a knowledge model database to interpret natural language queries by identifying entities and relationships, constructing structured queries, and executing them against a knowledge base to retrieve relevant FAQs and answers, leveraging Named Entity Recognition and Relationship Extraction techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If keyword-based search is used to retrieve information from a database, then the system is simple to operate, but precision and recall remain below 40% due to ambiguous words and different referring ways
Solution Approach 1:
The patent introduces an intermediary layer between the user's natural language query and the database search. This intermediary translates the query into structured assertions using a knowledge model, enabling more accurate retrieval while maintaining natural language input simplicity.
Solution Approach 2:
The system changes the parameter of query representation from simple keywords to structured assertions with semantic relationships. This transformation allows the system to capture the meaning and context of queries, significantly improving precision and recall while keeping the user interface simple.
2Measurement precision
If ontology-powered approaches are used to improve precision, then measurement precision improves, but device complexity increases due to formal syntax requirements
Solution Approach 1:
The patent segments the complex ontology processing into two distinct layers: a simple user interface that accepts natural language, and a backend translation mechanism that converts it to structured assertions. This segmentation allows high precision without exposing the complexity to users.
Solution Approach 2:
The system uses an intermediary translation layer that automatically converts natural language queries into structured assertions. This intermediary handles the complexity of ontology processing internally, allowing users to benefit from high precision without dealing with formal syntax requirements.
3Ease of operation
If natural language queries are processed without structured interpretation, then ease of operation is maintained, but recall is poor due to ambiguous words producing erroneous results
Solution Approach 1:
The system transforms the query processing parameter from direct keyword matching to structured assertion generation. This change enables the system to maintain natural language input simplicity while dramatically improving recall by capturing semantic relationships and disambiguating terms through the knowledge model.
Data Source
AI summary
A system and method for information retrieval are presented. A client computer receives a natural language query comprising an array of tokens. A query processing server analyzes the natural language query (interpreted as a question) to identify a plurality of terms and a relationship between one or more pairs of the terms according to a knowledge model defining interrelationships between a plurality of entities. A set of assertions is constructed using the relationship between the pair of terms, and a query is executed against a knowledge base of frequently asked questions, corresponding answers, documents and/or data using the set of assertions to generate a set of results. The knowledge base identifies a plurality of items, each of the plurality of items is associated with at least one annotation identifying at least one of the entities in the knowledge model. The set of results are transmitted to the client computer.