Design And Implementation Of Enterprise Relationship Extraction System Based On Deep Learning

With the rapid economic development and the gradual maturity of the business society,the development of enterprises is also facing huge opportunities and challenges.When enterprises make business decisions,they need more abundant and multi-dimensional information to provide support.As a part of enterprise resources,enterprise relationship has an important impact on an enterprise.It is the business data must be analyzed and mastered when making business decisions.These relationship networks can help enterprises evaluate market risk effectively,understand market dynamics comprehensively,reduce decision costs and improve decision efficiency.However,with the rapid growth of data in the Internet,the huge amount of data greatly reduces the efficiency of information acquisition.Therefore,how to mine and utilize the unstructured data through artificial intelligence technology and automatically obtain the potential enterprise relationship network is of great commercial and research value.In view of the above problems,this paper proposes an enterprise relationship extraction system based on deep learning.Enterprise relationship extraction system is a service platform designed for enterprise users,which can automatically mine enterprise relationship and build relationship network.It provides users with data management,relationship extraction,user management and other functions.Through the system,users can build their own enterprise relationship network efficiently which can provide information support for business decisions.In addition,in the function of relationship extraction,named entity recognition algorithm and relationship extraction algorithm are realized to recognition enterprise entities and extract enterprise relationships.Firstly,this paper analyzes the requirements of the enterprise relationship extraction system,divides the roles of users,determines the business requirements,and does a detailed analysis to the functional requirements and performance requirements.Secondly,according to requirement analysis,the overall design of the system from functional modules design,architecture selection,interface design and database design based on entity relationship graph.Then,describes the detailed design and implementation of each functional module.And the algorithm design and implementation of named entity recognition model and relationship extraction model.Finally,through the system test to verify whether the system meets the needs of users,including comparing the model to analyze its effectiveness and testing the function and performance of the system.