Applying Data Mining In The Analysis Of Monitor Of Mobile Communication Network

Reported by some equipments of mobile communication, Network Monitor List is some kind of information which mainly records the process of the error occurrence experienced by the equipments. Data mining refers to extracting interesting pattern or knowledge from large amounts of data. This thesis mainly discusses data mining and its application in the analysis of Monitor List of Mobile Communication Network and describes some approaches to find non-trivial, implicit, previously unknown and potentially useful information or patterns from Monitor Report data.Based on the analysis of the data mining technologies (association rules mining, the clustering ) and Mobile Communication Network Monitor, the primary work of this thesis includes: perform the data preprocessing of Mobile Communication Network Monitor, using technologies of data cleaning, data integration and Attribute-Oriented Induction;improve the Apriori algorithm using the Hash table and data partition and apply the improved algorithm to mine multidimensional association rules; search list from the database of Mobile Communication Network Monitor List based on the technology of important-words using both Euclidean distance and cosine distance to measure the similarity,.cluster the unstructured parts of the Monitor List using the fuzzy clustering.The experimental results of the applications about different data mining methods in Mobile Communication Network Monitor are given in this thesis, and what we get from the experiments are as follows: Mobile Communication Network Monitor rules could be found in analyzing them using data mining technologies; we could get the connotative relations between lists using the clustering and Cosine distance is more effective than Euclidean distance in the clustering and so on.