Research Data Mining Methods Based On Base Set And Concept Lattice

Data mining which is called Knowledge Discovery in Database is to extract or mine knowledge from large amounts of data. It involves the knowledge of lots of subjects and develops with these subjects. Association analysis applied in many fields is one of the important functions of data mining. Association analysis is to find association rules hidden in the data and the rules describe some interesting relations of the items in the given data set.Concept is method for human expresses their knowledge. The knowledge discover of database is a course of finding useful knowledge and forming the concepts. Concept lattice is an effective tool for concept discovery from data, having the ability to embody relationship of concepts in a vivid and concise way. Concept lattice has been widely used in many research fields. It is an important technique in the field of mining association rules.A few topics of data mining such as its basic concept, application, function, sorts, problems, etc. were dicussed at first in this paper. The content of association rules was discussed and a lot of algorithms such as Apriori and its transfiguration for mining association rules were introduced in the second. The traditional algorithms often suffered from the bottleneck of itemset generation because of difficultly confirming a suitable minimum support, In the other, the traditional algorithms often generated more redundant rules. Some research work was done to solve the above problems in this paper. The main content shows as follows:An algorithm for minging association rules based on base set was introduced. A base set which is generated by dynamical system propagating method based on seed items is a subset of the original dataset. The efficiecy of mining can be improved and the result rules are more interesting by using the base set instead of the original dataset.An algorithm for building simplified concept lattice which can be applied in the mining association rules was introduced. The simplified concept lattice was built from the large 1_itemsets by using the TIDs and support constraint. Two methods of finding the association rules based on the simplified concept lattice were proposed in this paper.An algorithm based on base set and simplified concept lattice were proposed to find association rules. The algorithm used the base set to build the simplified concept lattice and then mining rules from the lattice. In the end of this paper, the algorithm was applied in the field of spatial data mining of GIS and got a feasible result. The spatial data can be altered to the suitable form for mining association rules by using the technique named Spatial Join Index (SJI). Two methods were applied to get rid of the redundant and meanless rules in the paper.