Design And Implementation Of Configurable Knowledge Extraction System

With the rapid development of the knowledge economy,people have increased theirdemand for domain knowledge.The degree of people's access to knowledge clearly affects the quality and speed of their own development and development in the profession.Knowledge extraction methods can help people to obtain specific domain knowledge from a large amount of unstructured data,based on different backgrounds of users and solving multi-domain knowledge extraction,a configurable way of operation is required.Based on the above background,this paper designs and implements a set of software for extracting knowledge from unstructured data,which can better assist users in extracting domain knowledge in a more convenient way.This paper analyzes the process and actual requirements of knowledge extraction in detail.It designs the system based on the principle of improving the user's work efficiency and user-friendliness.It uses Spring Boot and Maven to build a microservice framework,and uses Postgre SQL database software and Java language to implement the background,and uses Angular JS and Bootstrap to implement the foreground functions,to design and implement configurable knowledge extraction system.The system is mainly designed and implemented with five functional modules.The data preprocessing module is used for the management of data sources and natural language processing.The data extraction module is a concrete process of knowledge extraction designed to implement entities extracting,relationships extracting,features extracting,weak supervision markers and statistical reasoning.The check feedback module is used to check and tag data for predictive data.The resource search module provides full-text retrieval of resources,and the statistical information module displays the changing trend of key indicators and iterative effects.Since different areas of knowledge extraction requires different processing processes and methods,the system satisfies the user's specific implementation through the drag-and-drop arrangement process and configuration knowledge extraction.This paper first analyzes the background,significance and research status of knowledge extraction.Then it analyzes the system from the use case analysis and data flow analysis,divides the module functions,and then designs the system architecture,database and core function implementation flow,and finally shows the system implementation page and tests the main function.The system has been released and tested on a small scale and the system is operating normally.The configurable interactive interface allows users to successfully completeknowledge extraction tasks according to their needs.It can be used not only by people with related professional backgrounds,but also for general users.