An Indexing Network Model for Information Services and Its Applications

Along with the enormous amount of information service resources on the Internet, it is increasingly necessary to consider the urgent problems it has brought, such as diversity, heterogeneity, disorder, and redundancy. Existing service technologies like P2P, grid computing, Web service, and cloud computing, mainly focus on a certain type of service resources, and they do not provide a basic model that can organize and manage various types of service resources existing on the current Internet. Therefore, to achieve better organization and management of service resources, thereby providing more valuable information services for Internet users, this paper introduces an indexing network model and five related normal forms to advance the field. As a basic model, indexing network organizes and manages various information service resources through analyzing the relationships among them. On this basis, applying the model to web page resources, we show two case studies. They are implemented using cloud distributed systems (Hadoop + Habse + Zookeeper) built on Sugon-Tongji cloud platform located at Tongji University. 70 million web pages are crawled and two indexing networks are constructed. Through analysis of service applications provided by them, we show the advantages of the proposed indexing network model to organize web pages and keywords, which can provide and support more knowledgeable and valuable search related services, thereby better meeting the complex and diverse needs of Internet users.