The Research And Application On Mining Association Rules In Multidimensional Data

With the commercialization process of data mining technology, the value of data received more and more attention. As a column of New York Times says that "big data" era is coming, decisions will be increasingly based on data and analysis, which is not based on experience and intuition. Data as new means of production,will produce a new economic operation mode.For solving the dispute in rate survey, association rule technology was studied in this paper. Analyzed the data of audience with association rule algorithm, we expect to find out the potential rules of program transmission,and generate a new decision making system for radio station.The design ideas and improvement methods of Apriori series algorithm are analyzed in depth. An improved association rules algorithm which based on matrix was proposed. The algorithm is pruned by matrix reduction. Simulation experiments contrast the improved algorithm with Apriori and AprioriTID algorithms. The results show that the improved algorithm improves the efficiency by reducing the matrix and using the "or" operation instead of summing the support. As a result that the data of Long Guang's audience serves as data cube,so the combination of OLAP technology and the improved algorithm based on matrix can better serve the decision making. In addition, aiming at the limitation of radio transmission technology, the writer sets up the audience database system of HLJ People Radio Broadcasting Station to record the interaction of the audience and the program. The system uses OLAP technology to show the managers of radio interactive information from multidimensional angle, and provides an effective way to prepare the analysis data for data mining.