Knowledge Q&A Retrieval Across Structured Data and Documents
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for searching business knowledge require users to manually retrieve and combine structured data and unstructured document knowledge, leading to low efficiency and accuracy in acquiring knowledge.
Innovation Solution
A knowledge question and answer method utilizing a machine learning model to retrieve target data fields and knowledge documents from separate knowledge bases, determining answers based on metadata and business knowledge documents, and displaying the results directly to the user.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If users manually retrieve and combine structured data and unstructured document knowledge, then they can access both types of knowledge, but the efficiency and accuracy of knowledge acquisition deteriorates
Solution Approach 1:
The system performs self-service by automatically retrieving structured data from data tables and unstructured knowledge from documents, then combining them without requiring user intervention. The machine learning model autonomously processes the integration of different knowledge types, eliminating manual retrieval operations.
Solution Approach 2:
A machine learning model acts as an intermediary between the user's knowledge needs and the multiple knowledge sources. The model receives user questions, automatically queries both structured data tables and unstructured documents, synthesizes the information, and presents integrated results, serving as a mediator that handles the complexity of multi-source knowledge retrieval.
2Loss of information
If users query field data in data tables and retrieve related documents separately, then comprehensive knowledge can be obtained, but the operation complexity increases
Solution Approach 1:
The system merges multiple knowledge retrieval operations into a single unified process. Instead of requiring separate queries for data tables and documents, the machine learning model simultaneously accesses both structured data and unstructured knowledge sources, combining them into integrated results that provide comprehensive knowledge in one operation.
Solution Approach 2:
The machine learning model serves multiple functions: it acts as a natural language processor, a database query generator, a document search engine, and an information synthesizer. This multi-functional approach allows the system to handle diverse knowledge retrieval needs through a single interface, improving ease of operation while maintaining completeness.
3Device complexity
If a unified knowledge base is used to store both structured and unstructured knowledge, then system complexity is reduced, but retrieval precision deteriorates
Solution Approach 1:
The system maintains segmentation by keeping structured data tables and unstructured documents in separate knowledge sources, preserving their distinct characteristics and optimization for different query types. The machine learning model intelligently routes queries to appropriate sources or combines results from both, maintaining retrieval precision while managing complexity through intelligent orchestration rather than physical unification.
Data Source
AI summary
The present disclosure relates to a knowledge question and answer method, a computer-readable medium and an electronic device, the method includes: acquiring a target question input by a user in a natural language; retrieving, by a machine learning model, in a first knowledge base according to the target question to obtain a target data field for answering the target question, and retrieving in a second knowledge base according to the target question to obtain a target knowledge document for answering the target question, and determining a target answer to the target question according to the target data field and the target knowledge document, where the first knowledge base is configured to store metadata fields of a business data table composed of business data, and the second knowledge base is configured to store business knowledge documents; and displaying the target answer to the user.