Research On ANP Technique For Web Resource Quality Assessment

With the rapid development and popularization of Internet, more and more people obtain information from Web. Because of its extensive sources of data, Internet is of dynamics, independence, and diversity. Due to the huge amount of heterogeneous information on the Web, information overload and pollution are big issues. The Web resources quality evaluation is essential and helps to acquire and use information correctly. The goal of Web resource evaluation is how to obtain high quality data from Web.This paper introduces a system about Web source evaluation where Analytic Network Process (ANP) is used to analyze Web resource. Firstly we survey the related work and study the evaluation criteria. Different evaluation models are compared. Aiming at different applications, the thesis focuses on two kinds of evaluation models. In one model, this paper selects multidimensional and multilayer web quality model-WebQM. In the other model, the characteristics of websites are evaluated by users. Three big criteria blocks (information content, arrangement and design, performance for easy use) consist of 12 evaluation criteria. They are built and explained in detail. Secondly, based on the model data is obtained as the parameters, which are measured by experts. ANP is used to integrate the sub-results. Finally, the analysis of sensitivity is carried out to check the rationality of the system.We can get an overall target of quantization while evaluating by using ANP. It is good to compare a lot of the criteria, so the inaccurate qualitative evaluation can be avoided. By using such a method reasonable results can be achieved. The suitable web sources can be selected by users.