Web Mining Research And Applications

This thesis mainly discusses Web information mining technology, systematically expounds the Theory and method of web data mining, it intends to apply the data mining technology to find structured information on special domain, and maps these data into relational data model then stores in the realational database. So it can reach the goal of using Relational Database Management System(RDBMS) to manage and query web information source, and uses user access mode mining and favorite subjects discovery technology to accurate position necessary information for user.This thesis briefly introduces the background of Web information mining and application value, and put forward the Web data mining system overall objectives and the overall structure, and points out the system implementation technical route.The paper presents the design of a Web document model, marks the nest level structure relations Web page into a tree map symbol through HTML Web page modeling, designs and implements the algorithm of domin information block finding in web pages with a group of inspiration rules.The thesis uses Web metadata extraction technique to extract domain information, firstly models the web page with defined data model (OEM model), and extracts the web metadata with web page structured mining algorithm. And based on the STORED mapping relationship related techniques and methods, turns the semi-structured data into relational data.The thesis discusses user access mode mining and favorite subjects'discovery technology, and makes a comparative analysis between the user access mode mining and related rule mining, points out their differences and similarities, according to the characteristics of access route, gives the corresponding mining method.