Huang Wei

Publications (2)0 Total impact

• Article: A Framework for an Ontology-based E-commerce Product Information Retrieval System

  Zhang Liyi, Zhu Mingzhu, Huang Wei
  [show abstract] [hide abstract]
  ABSTRACT: With the rapid development of e-commerce, online shopping has become an important part in people's lives, in order to support the smooth development of e-commerce activities, how to provide users with an efficient and practical product information search method has become an urgent and critical problem. This paper presents a framework for an ontology-based e-commerce product information retrieval system and proposes an ontology-based adaptation of the classical Vector Space Model with the consideration of the weight of product attribute. A computer and components related ontology has been built, which is adopted to annotate the html documents and construct concept vectors of the documents. Then the system test is done and the experimental result indicates that our proposal is better than the traditional keywords based search.
  Journal of Computers. 01/2009;
• Source

  Available from: Liyi Zhang

  Article: Semantic Focused Crawling for Retrieving E-Commerce Information

  Huang Wei, Zhang Liyi, Zhang Jidong, Zhu Mingzhu
  [show abstract] [hide abstract]
  ABSTRACT: Focused crawling is proposed to selectively seek out pages that are relevant to a predefined set of topics without downloading all pages of the Web. With the rapid growth of the E-commerce, how to discovery the specific information such as about buyer, seller and products etc. adapting for the online business user becomes a focused issue to the information search engine. We present a novel semantic approach for building an intelligent focused crawler which deals with evaluating the page’s content relevance to the E-commerce topic by the domain ontology and the hyperlinks connection to the commercial web pages by link analysis. In the process of crawling, the domain ontology can evolve automatically by machine learning based on the statistics and rules. Experiments have been performed, and the results show that our approach is more effective than the other traditional crawling algorithms, and prevents the topic-drift with higher harvest rate.
  Journal of Software. 01/2009;