Research On Website Classification Based On Linking Open Schema

Website classification is the process of specifying the existing classification tag to some specific website. To classify the content of the web page is an significant method and is essential to many tasks in the field of information retrieval; it can not only allow users to get more effective information and improve the efficiency of search, but also it can largely alleviate the problem of information chaos. The main content of this thesis includes:1. Construct the website category system based on zhishi:schema, a project based on linking open schema and get the feature word list of each category from three knowledge base as Zhishi.me, Babel.Net and cilin of Harbin Industry College, all above is the basis of website classification.2. Put forward a method of constructing feature word list, which can represent website feature with tag content inside website homepage and tag content of link pages of homepage.3. Design a weighted matching algorithm between the website feature word list and the possible feature word list of each category in website classification, and a website classification algorithm based on the maximum matching degree.4. Through the experiment, the validation of the matching algorithm and the website classification algorithm that is put forward is testified.5. Based on the website classification algorithm, the website recommendation algorithm is achieved and a website navigation system is designed with two functions as website classification and website recommendation.The main contributions of this thesis are proposing a website classification algorithm based on LOS, applying the concept of tag content and neighbor web page to the website classification, and using multiple knowledge bases to get the feature of classification text.