Association rules and fuzzy association rules to find new query terms.
Miguel Delgado‐Rodríguez, Marı́a J. Martı́n-Bautista, Daniel Sánchez, José-Marı́a Serrano, M.A. Vila · 2003
We present an application of association rules and fuzzy association rules to find new terms that help the user to search in the web. A first query is made to the web and an initial set of documents is retrieved. Considering the terms of the collection as items, text transactions are constructed and association rules are extracted. Fuzzy association rules are also considered when the presence of an item in a transaction is a weight between zero and one. A selection of the best rules is carried out and the terms in those rules are offered to the user to refine the query.