An automatic classification technique and tool for information retrieval of web documents

In this paper we describe a technique, and its prototypical implementation, for Information Retrieval of text documents and Web pages and sites. The technique is based on automatic classification methods, more precisely on an unsupervised and hierarchical clustering method. The defined technique produces a hierarchical classification tree from the set of analyzed documents, and therefore classifies them with respect to paradigmatic similarity relationships. This technique has been implemented and a prototype tool has been produced, completely realized in Java. This prototype provides the user with information retrieval functionalities, by means of a search document map based graphic interface. The main functionalities of the tool are described in this paper.