Study On Tourism Emergency Association Rule Minging Algorithm

With the continual growth of national economic, tourism has developed rapidly and made more and more contributions to economic development. However, emergencies occurred in tourism industry have obstructed its development and brought huge loss to tourists.This thesis is based on National Natural Science Foundation of China(60773112): information pattern analyzing and forecasting study based on tourism emergencies; Beijing Natural Science Foundation (4082021): tourism emergencies data mining and intelligent prediction of tourism emergencies. In this thesis, combining data mining with tourism emergencies, we obtain the useful information from database. The work being done is as follows:(1) Do preprocessing of tourism emergency database. Abstract attributes from tourism emergency database. Through data cleaning, data integration and transformation, we get the database which is available.(2) Present the concept of the attribute item equivalence class based on frequent pattern tree and build equivalence class to mine frequent pattern item in frequent pattern tree. Comparing with traditional FP-tree, we improve the algorithm efficiency a lot.(3) Bring forward a new association rule algorithm which is obtaining frequent patterns based on equivalence class, describe the algorithm in detail and analyze the time and space complexity qualitatively. We apply this algorithm in tourism emergency and get the data mining rules hidden in the database and validate its feasibility. We compare FPBE and FP-tree and indicate the efficiency of FPBE.We improve the traditional frequent pattern tree algorithm based on the previous research and bring forward a new association rule mining algorithm and apply the new algorithm in tourism emergency. The research results extend the scope of application of association rules and provide support to tourism emergency prediction system.