A fuzzy linguistic model to evaluate the quality of Web sites that store XML documents

The development of tools to find quality information on the Web is currently a pressing need. The aim of this paper is to present an evaluation model based on fuzzy computing with words to measure the information quality of Web sites that store XML (eXtensible Markup Language) documents. This model evaluates the information quality of Web sites using only users' perceptions, and therefore it is user-centered. Fuzzy linguistic techniques are involved in the quality evaluation process to create a user-friendly framework. This model is composed of two main components, an evaluation scheme to analyze the information quality of Web sites and a computing method of quality ratings of Web sites. The evaluation scheme presents both technical criteria related to the Web site characteristics, and criteria related to the content of XML documents stored in the Web sites. The quality ratings represent the ability of Web sites to meet user requirements. Linguistic quality ratings are obtained by combining linguistic evaluation judgements provided by Web visitors on the different evaluation criteria. The computing method is based on two operators for fuzzy computing with words, the LOWA (Linguistic Ordered Weighted Averaging) operator and the LWA (Linguistic Weighted Averaging) operator. The later allows to manage relative importance degrees among quality criteria in the evaluation process. This model uses the power of XML Schema language to improve the representation of documents in the Web with semantic characteristics related to their quality and thus it is useful to search quality resources in XML format. Web site quality ratings could be used by Web retrieval systems to help users to find the highest quality XML resources for their information needs. Additionally, this model could be helpful to Web developers to improve the quality of Web sites from a user's point of view. (C) 2007 Elsevier Inc. All rights reserved.