179 Chapter 19 WDM-WISA: a metadata based tool for Web information analysis William W. Song, SISU, Electrum 212, 164 40 Kista, Sweden, william@sisu.se Anders Segerberg, SISU, Electrum 212, 164 40 Kista, Sweden, seger@sisu.se Abstract Metadata has been considered to be "additional data to describe web data". There are many approaches to construct these additional data. The objective is to provide a better, effective description of the web information for effective and efficient analysis and management. Through investigation of a number of the metadata methods, we consider it indispensable to develop a simple metadata model and use the model to realize an effective metadata based tool for web information analysis. In this paper, we present a metadata model, WDM - Web Document Model, for description of web resources. We focus on an approach to web information analysis and management based on the web information metadata model. A tool is developed based on the metadata model WDM. The tool attempts to be able to take in web information, no matter in which metadata model it is described, and produce a uniform structure for the information. The tool is also expected to edit and analyze the information. It can yet be used as a metadata based web page creator as well. Keywords. Web resources, Metadata, Web document analysis, Resource Description Framework 1 Introduction The use of the World Wide Web (web) has brought people a great convenience of obtaining information quickly and easily. However, the ever-growing information on the web has meanwhile brought people a great difficulty in quickly and easily finding the exact information they need. A problem causing the difficulty is that web information and resources are less well structured. Most of web resources (web pages, web documents) contain little additional information used for identifying the web item. Now, the problem is receiving increasing attention from both researchers and the end users. The term "metadata" is introduced to indicate this additional information to web information items for the purpose of description and identification. For example, from a pile of web documents searched, we hope to quickly and effectively find what we exactly need, analyze its properties, and then classify it in a certain category. We also hope from one document we can dig out which piece of information that we need. All this requires a framework, which contains some data to support to identify web documents and document items. The framework is so called "metadata". Metadata can be considered to be the particular part of a web resource, which contains information intentionally and specifically for the web resource identification. There are many metadata models at present to support metadata information to web resources. Some well-known metadata models, among others, are Dublin Core (DC) [1], Meta Content Framework (MCF) [2], and Resource Description Framework (RDF) [3]. These metadata models are usually focusing on a particular domain. For instance, DC is designed for the description of publications in libraries. However, we consider it important to define an integrated metadata model, which is able to integrate the merits of various metadata models. In other words, a general-used metadata model is needed to interpret and represent the other metadata models developed for special purposes. We also maintain that a good metadata model should meet these requirements: simple in presentation and modeling, rich in semantics and expressiveness, easy to use, and flexible to adapt to various local models (such as relational