Research On Key Issues In Construction And Application Of Fuzzy Concept Lattice

As a powerful tool for data analysis,the concept lattice has been widely used in many fields such as data mining,knowledge representation,machine learning,information retrieval,pattern recognition,expert system,artificial intelligence and so on.In the real world,the information is usually vague or uncertain.Fuzzy concept lattice is the combination of fuzzy theory and concept lattice,which can be used to analyze and process fuzzy information.The construction of fuzzy concept lattice is an NP-complete problem.In the worst case,with the increase of the scale of formal context,running time of fuzzy concept lattice construction increases exponentially.The construction of fuzzy concept lattice is the premise of its application,and it is also a bottleneck problem.In this paper,the construction algorithms of fuzzy concept lattice are studied and improved.Concept lattice and fuzzy concept lattice play an important role in many applications such as intelligent disease diagnosis,association rule mining and personalized recommendation service.In this thesis,some key issues in the construction and application of fuzzy concept lattices are studied.The main work of this paper is as follows:(1)The construction algorithms of fuzzy concept lattice are studied and improved.The existing incremental construction algorithm of precise concept lattice is extended.The operation method of fuzzy sets is added,which can be applied to fuzzy concept lattice.The pruning technique is added to improve the incremental construction algorithm of fuzzy concept lattice based on object.According to the dual property of fuzzy concept lattice,the incremental construction algorithm of fuzzy concept lattice based on attribute is designed.At the same time,an algorithm for inserting objects in the fuzzy concept lattice based on attribute is proposed.The experimental results show that the proposed algorithm with pruning steps reduces the running time,and improves the construction efficiency of fuzzy concept lattice.(2)A novel intelligent disease diagnosis method based on fuzzy concept lattice is proposed.Symptoms and the corresponding extents(e.g.,frequency,severity,and duration)of each disease can be extracted to form a fuzzy concept lattice.The fuzzy concept lattice of the symptoms and their extents to be diagnosed needs to be constructed to match the fuzzy concept lattice of possible diseases.The similarity between the above two types of fuzzy concept lattices can be calculated and used to aid for effective diagnosis.Naturally,the disease with the largest similarity is the finding of intelligent diagnosis.(3)A novel personalized recommendation system based on concept lattice is designed,which consists of the offline part and the online part.In the offline part,the formal context and the concept lattice are constructed from the transaction database,and the association rules based on concept lattice are extracted and stored in the rule library.The new added data are used to update the concept lattice and the rule library regularly.The online part uses a hybrid recommendation strategy,which combines the recommendation algorithm based on association rules and collaborative filtering recommendation algorithm based on concept lattice.In the online part,the behavior data of target user,the concept lattice and the rule library are used to calculate the ordered recommendation results,which are returned to the user.(4)An algorithm for mining and updating association rules based on fuzzy concept lattice is proposed.When a new attribute is added into the fuzzy concept lattice,it is not necessary to calculate all the frequent nodes and association rules.According to the incremental construction algorithm of fuzzy concept lattice,it is only necessary to deal with the new nodes that have changed.Therefore,the amount of calculation is reduced.The incremental construction algorithm of fuzzy concept lattice based on attribute is extended.The steps for generating and updating association rules are added.According to the extended algorithm,the fuzzy concept lattice can be constructed,and the corresponding association rules can be generated and updated at the same time.