Captain nemo: A metasearch engine with personalized hierarchical search space

Personalization of search has gained a lot of publicity the last years. Personalization features in search and metasearch engines are a follow-up to the research done. On the other hand, text categorization methods have been successfully applied to document collections. Specifically, text categorization methods can support the task of classifying Web content in thematic hierarchies. Combining these two research fields, we have developed Captain Nemo, a fully-functional metasearch engine with personalized hierarchical search spaces. Captain Nemo, retrieves and presents search results according to personalized retrieval models and presentation styles. Here, we present the hierarchical Web page classification approach newly adopted. Captain Nemo Jets users define a hierarchy of topics of interest. Search results are automatically classified into the hierarchy, exploiting hierarchical k-Nearest Neighbor classification techniques. The user study conducted demonstrates the effectiveness of our metasearch engine