Based On Clustering Algorithm For Sequential Pattern Mining And Its Application In The Quality Traffic

In recent years, with the rapid development of wireless communication technology, the telecommunication business increase unceasingly, the penetration rate of Smartphone increased rapidly,3G,4G applications appear continuously, that make the business data in telecommunication industry become more and more, so telecommunication industry becomes a typical data-intensive industries. Because people contact each one every day through communication tools such as mobile phone, so there are producing large amounts of voice services data every day, The voice services data accumulated over a long period of time become an important resource and wealth, how to find valuable information from these massive amounts of voice services data is necessary.Data mining is a kind of tool which can mining valuable knowledge from huge amounts, Its function is finding useful and hidden information from a large amount of data by the algorithms. Among them, the sequence pattern mining and clustering mining are important branches of data mining, they have been applied in many fields, for example, customer purchasing behavior analysis, fraud detection, network intrusion detection, etc. On the basis of consulting a large number of domestic and foreign literatures, this article adopts the sequence pattern mining and clustering analysis to mining the voice services, at last to obtain valuable information and to predict telecommunication business decisions or action.Based on the voice services data has a huge amount of data, this paper did some research on the partial data object, made a detailed analysis of the data’s attribute, analyzed the reasons which influence the quality of the current telecommunication, combined with K-means clustering algorithm and the improved prefixSpan sequential patterns mining algorithm to do cross mining, built a mining model and analyze it. In this paper, the main research work is as follows:1. Proposed the sequence pattern mining method for voice telecommunication services.2. Designed and implemented a sequential patterns mining model on voice telecommunication services about user satisfaction.3. The customer satisfaction was divided into "excellent, good, medium, general and poor" five grades, according to the telecommunications business "Real-time business、 service accessibility、business quality、business keeping、business integrity" clustering in five aspects, respectively is designed and implemented the model and its related algorithm.4.Improve the PrefixSpan sequential pattern mining algorithm, obtained the Sequential pattern mining results through the sequential pattern mining for the clustering results.5.Based on the telecommunication business data sets, had done the experiment by the improved PrefixSpan algorithm and the Apriori algorithm for mining sequential patterns, the experimental results shown that the improved PrefixSpan algorithm has better performance.Results of this study obtained the corresponding sequence mode, provided a basis for decision making for the telecommunications, and provided better voice calls.