Document clustering based on concept lattice

There is in our days an increasing interest on application of concept lattices for data mining, for representation of the concept hierarchy generated from the underlying data set. The paper represents a document query system that clusters the documents into groups using a generated concept lattice In the first phase. The users can start the query by entering a set of keywords. The system return the concepts closest to the query vector. The users get a list of neighbouring concepts too, thus they can select the way to navigate to reach the result document set. The final result is produced using an iterative relevance refinement feedback method.