Design And Implementation Of Personalized Service Subsystem Of Color Ring Back Tone Professional Analyses System

With the rapid development of information technology, growth of information carried on the internet platform follows a exponential trend. Although the information explosion greatly enriched people’s life, it will make it more difficult for people to get contents which they feel interested in. Whether for the E-commerce platform or telecom operators, the personalized service demand aiming at user’s level has become more and more urgent.As the most popular value-added telecommunication services, the business volume of color ring back tone services has tended to be saturated, and the further development may face great challenges. There are some problems exist in the color ring back tone services operation process, such as the plan lacks of data support, not fully analyze user’s information and usage habit, and the depth in operation is insufficient, etc. Aiming at the above key problems, the color ring back tone professional analyses system of Zhejiang province separate the contents like operation analysis, data mining algorithms and marketing execution from business platform, combined personalized services system among them, and provide data and system support for accurate marketing by user behavior analysis and data mining algorithms.The personalized service subsystem designed and implemented by the project is based on the color ring back tone professional analyses system of Zhejiang province. Its main purpose is to continue to promote color ring back tone business development and enhance business income. The system can filter, integrate and estimate the users’attribute information and interbehavior information, and use personalized recommendation service to enhance marketing accuracy and success rate. The project will establish its own scoring mechanism of unique users and color ring back tone, and design as well as implement personalized service subsystem finally by comparing the advantages and disadvantages of various recommendation algorithms, summarizing the recommendation methods and characteristics of all kinds of personalized platform and fully analyzing the demand and function boundary of the color ring back tone professional analyses system of Zhejiang province.The core data mining algorithm of the system is collaborative filtering algorithm. According to the main problem faced by data mining algorithm, a method based on user clustering analysis is applied to collaborative filtering algorithms in the paper, which will limit "nearest neighbors" searching object in the most similar cluster, thus to ease matrix sparsity. Since the similarity of internal users in each cluster is higher, and the interest is relatively concentrated, it will avoid the effect caused by larger data deviation to recommendation result, thus to enhance the recommendation accuracy as well as the implementation efficiency of recommendation algorithm, to reduce computational matrix dimension and to enhance the system scalability.Paper is organized as follows:The first chapter briefly describes the research background, significance and applied value of this study. The second chapter describes the technical background on relevant topics, including the whole network structure and business requirement of the color ring back tone professional analyses system as well as the introduction of core algorithm related to personalized service subsystem. The third chapter mainly describes the whole design plan of personalized service subsystem as well as clustering optimization method of collaborative filtering algorithms. The fourth chapter mainly describes the detail design of personalized service subsystem, including the design and implementation play of various main functional modules. The fifth chapter includes analysis of each system tests. The sixth chapter of the paper summarizes the work, combined with the current business situation, and proposes further research in this field of work.