Research And Design Of Drug Procurement Recommendation Information Management System Based On Decision Tree Model

In the Internet era of rapid development,huge amounts of data received by user at the same time,users are more expectations for more accurate information from the huge amounts of data,on the basis of the big data,machine learning,data mining and other emerging concept was put forward.In which a large number of users in daily life,such as in the process of online shopping is the most commonly used to recommend system,based on the history of the user purchase data,analysis of purchasing behavior,get buy the goods of high expectations,pushed to the user.Increase the user's purchasing power and promote the increase of turnover.But in front of the huge demand,the current user terminal retail enterprises lack a complete set of purchase decision-making recommendations on the decision-makers to provide help for the company,so that a large number of terminal retail enterprises only rely on historical experience for purchasing decisions,which affect the benefit of the company.In this context,three aspects of research work will be carried out:data integration and analysis,decision tree model construction and recommendation system integration.Data integration and analysis,mainly aimed at the pharmacy commodity base integrating data and sales data,and integration of them construct the decision tree necessary attribute values,through the analysis of the attribute value for data aggregation,get the attribute value if I purchase a metamodel,paved for constructing decision tree.Decision tree model structure on research to understand the decision tree algorithm,and through the experimental data for decision tree construction algorithm to improve,improve its extreme cases can't choose the root node,the point for the research of the core of the whole paper.Recommendation system integration is mainly to integrate the data collection and decision tree,the importance of attributes is obtained by decision tree is returned again after to integrate data collection to extract data,extracting the data for decision makers to offer help.The integration and analysis of data is the basis of this topic,and its research significance lies in how to provide data support for the system through data integration and analysis.In this paper to construct the decision tree algorithm to study and improvement of paper core points,based on the concept and theory of this algorithm,in extreme cases the root node selection decision tree model problem put forward the improvement measures and ways of the author.As the final design concept of the thesis,the recommendation system integrates the data set and decision tree to form the information management system based on the decision tree model.This article from the data integration and analysis of the advantage,choose the ID3 algorithm of decision tree constructed by topic in aspects of the drug purchase recommendations based on decision tree model subsystem,the information management system for step of the research process.In the core research point,the algorithm of the decision tree construction model is adopted to improve and contrast the situation of the root node in the extreme cases,and the improved algorithm is introduced with the weighted thought.Through the improved decision tree model construction algorithm and the original construction algorithm,the decision tree construction process is compared.