Research On The Decision Support System Model Based On Web Mining

Accompanying the development of information technology and E-commerce, the model of knowledge and decision support under the network has been becoming a new hotspot topic in the domains of E-commerce, management science and artificial intelligence etc. In the information society with the network economy developing rapidly, it is very important both in science and in application to make a thorough study on the models of knowledge discovery and decision support based on the web mining. The main contents discussed in this dissertation are as follows:By referencing to existing literature in related fields, the process, methods, functions, the missions and targets in each phase of the process model of data mining are summarized and reviewed, the concept, category and main tasks of web mining are expounded.The web documents classification model based on concept feature vectors are built. In order to manage and classify the semi-structure web documents, it advances the concept of concept feature vectors, describes the collection of concept feature words and the calculation method of weights, constructs the NB and Association Rules classifiers based on concept feature vectors, and verifies the validity and correctness of the web documents classification model based on concept feature vectors by experiments.The models of knowledge discovery based on the management of web usage data are set up. It expounds the URL-UserID associated matrix express method for browsing behaviors of web users, establishes the models of web pages clustering and web consumers clustering based on URL-UserID associated matrix, constructs the multi-Markov chains prediction model of web user browsing behaviors based on clustered web users, and the method of the mining model of maximal frequent sequences based on clustered web users and suffix trees.The models of knowledge discovery under the environment of semantic web are discussed. It introduces background, the system structure of the semantic web and web ontology, elaborates the expression model of domain ontology, discusses the classification process of web pages under the semanticweb, builds the classification model of web pages based on the domain ontology and the web usage knowledge discovery model based on semantic user profiles.The intelligent decision system model of multi-agent based on web mining is analyzed and constructed. It mainly discusses the system structure of the intelligent decision system under the environment of the semantic web, the assignments and goals of each agent system, the communication mechanism between different agents, and the key techniques for the implementation of the intelligent decision system based web mining.It discusses the application of the model of multi agent intelligent decision system based on web mining in the intelligent electronic learning system in the last chapter. It mainly analyses the decision function models of the electronic learning system, introduces the expressions of knowledge structures, and the designs of correlative databases, analyses the system structure of intelligent electronic learning decision support system and tasks of its each subsystem.