The goal of this chapter is to propose a methodology and tools to enhance information retrieval and navigation on the Web through contextual and conceptual help. This methodology provides users with an extended navigation space by adding a conceptual and a semantic layer above Web data. The conceptual layer is made of Galois lattices which cluster Web pages into concepts according to their common features (in particular their textual content). These lattices represent the Global Conceptual

more »

... xt of Web pages. An additional navigation layer is provided by ontologies which are connected to the conceptual level through specific concepts of the lattices. Users may navigate transparently within each of these three layers and go from one to another very easily. However, the navigation within Galois lattices may be difficult as the number of concepts grows very fast with the number of Web pages. The second contribution of this chapter consists in providing tools to help users navigate within a complex conceptual lattice. A new similarity measure is proposed to find the most relevant concept to start a navigation or to choose the most relevant concept to visit from a given navigation point. This similarity measure is based on Jiang and Conrath's measure used for ontology matching, extended to reflect conceptual information. This chapter illustrates these methodology and tools for Web information retrieval and navigation through example experimentations and presents future research directions-visualization in particular. 2