Esophageal And Gastric Cardia Pathological Features Associated With Analysis

Data Mining is the nontrivial process of identifying valid ,novel, potentially useful, and ultimately understandable patterns in data. It can be applied in all fields where data were copiously accumulated. After several decades mass survey of Esophagus Cancer and Cardiac Cancer, rich number of data have been accumulated, in which Data Mining technology can be applied to dig the high risk factors from those confused and disorderly pathological characters. In this paper, the Data Mining technique is applied in survey data of Esophagus Cancer and Cardiac Cancer for mining association rules and max frequent itemsets.After discussion and analyze the association rules mining techniques, we focus on the data preprocessing step, which is very important to the whole data mining process. First, Esophagus Cancer and Cardiac Cancer mass survey data are convered from text file to records in database, and then , changing multi-value attribute into Boolean attribute, filling in the missing value and deleting irrelevant attribute, are done. After data preprocessing, the form of data is fit for association rules mining and max frequent itemsets mining. Some interesting rules and frequent itemsets are mined that can be the guide for Esophagus Cancer and Cardiac Cancer mass survey.In this paper, the prototype system named CharacterMiner for mining association rules in Esophagus Cancer and Cardiac Cancer mass survey data is implemented, which is composed of data extracting, data preprocessing, association rule mining and results displaying.Through this study, certain experience of data mining technology application in Esophagus Cancer and Cardiac Cancer mass survey is accumulated. Appropriate analyze model is build up, providing information and modern method for Esophagus Cancer and Cardiac Cancer research work and mass survey.