Approximation Reduction Of Context And Its Application In Information Clustering

With the rapid development of network technology, the web resources present the volatile growth. How to find out the useful information most quickly and exactly has become a difficult problem of the retrieval technology.Face to the challenge, the data mining technology and knowledge discover technology emerged as the times require. As a powerful tool for the data mining, the concept lattice has been widely and successfully used in the field of information clustering. However, because the concept lattice is a complete lattice, the time and space complexity of building concept lattice presents an exponential growth with the size of the context. It has been one of the main obstacles that restrict it from being applied. So, in order to raise the building efficiency, many researchers studied the reduction of context.This paper studies the context and it's reduction, puts forward a context approximation reduction algorithm based on the Objects-Match. CROM algorithm which use the definition of the Objects-Match calculates the object-match between each object and the customers request. Then, the algorithm carries on the approximation reduction for the original context according to the limit of the OM. The limit of the OM controls the scales of the compatible subcontext. Constructing concept lattices which use the born compatible subcontext raises the efficiency of the concept lattices constructing not only, and assures the minimal losing, raises the serviceability of the concept lattices.The main contents of the dissertation includes,(1) Brief introduction of fundamental notions including definitions relevant to concept lattices and context reduction.(2) Research the sorts of the many-valued context, and choosing the conversion methods of the many-valued context to the single-valued context according to the classification. The conversion offers the viable ways for the approximation reduction of the many-valued context.(3) Research the methods of the standard reduction and the approximation reduction for the context. According to the disadvantages of the presented algorithm, the paper puts forward a context approximation reduction algorithm based on the Objects-Match, and designs a function to judge the algorithm. The experiment proved it control the size of the concept lattice not only,and raise the validity of the concept lattices.(4) Introduction the model for the information clustering based on the formal concept analysis, and the effect of the CROM algorithm.