Research On Key Technologies Of Enterprise Intelligent Decision OA System

With the development of information technology,relying on information technology,accelerating enterprise information construction,and further improving office quality and efficiency are the needs of enterprises to adapt to new forms of development.Achieve enterprise office automation reflects to a certain extent the modern management level of an enterprise.At present,the office form of most enterprises is mainly based on the OA(Office Automation System)system,so the degree of construction of the system is highly correlated with the efficiency of the enterprise and the level of enterprise management.In the process of enterprise development,daily office work and engineering construction through OA makes the data of each business system increase and become complicated.How to improve OA construction,effectively process,mine,and analyze these huge data will help realize the intelligent decision-making of enterprises,which will affect the future development of the company to some extent.For the OA system,more and more impractical function modules are integrated and it is difficult to exchange data with other business systems of the enterprise,and the decision-making performance is low.This paper will discusses the methods of classification and processing of information data in the system and the means of using these data for decision analysis,and put forward the basic architecture design of intelligent decision OA system.The research work includes the following aspects:1.This paper optimize information coding in the design of the system knowledge base module,in the aspect of effective organization and management of knowledge information,this paper proposes a method to improve the classification of massive information by combining LINQ technology when optimizing information coding.Compared with the traditional iterative query method of classification structure information,the new method greatly improves the retrieval efficiency of information and optimizes the organization structure of knowledge base information.2.For the data generated in the enterprise engineering construction,it is applied to the decision-making to generate the decision tree model for decision analysis.Aiming at the problem that the processing of continuous value attribute has high time complexity when dealing with massive data in the traditional C4.5 decision method,a threshold segmentation optimization method is proposed,which reduces the discretization threshold segmentation of continuous attributes in data.The calculation of points improves the efficiency of decision tree generation.At the same time,the correlation coefficient between attributes in probability theory(Pearson)is used to reduce the attributes in the data set.Combining the information gain rate of the attribute,the optimal subset of the decision attribute is retained,and there is no redundant attribute in the attribute subset,which improves the accuracy of the decision tree generation.3.Based on the requirements of fine management of OA system module functions and aiming at the intergration of different development methods,the paper adopts the idea of SOA(Service-Oriented Architecture)architecture,it builds the knowledge base module and intelligent decision module as the core components,and provides a large number of secondary development interfaces.By highly integrating other business systems of the enterprise,it becomes a streamlined and scalable enterprise portal platform.