Research on Olympics-oriented Mobile Game News Ordering System

Research on Olympics-oriented Mobile Game News Ordering System Yonggui Yang, Lei Li CISTR, Beijing University of Posts and Telecommunications, 100876 E-mail: ygyang410@sohu.com lilei@nlu.caai.cn Abstract. The Olympics-oriented mobile game news ordering system introduced in the paper involved the technology of natural language processing, machine learning and information extraction. We collected text information from the major Internet media websites about the table-tennis and badminton games of the 2004 Athens Olympics; then we transformed these texts information into structured data through information extraction. Based on this source data-base, we can provide game news to the end users in the form that they can read them on their mobile phones easily. Keywords: machine learning, information extraction, Short Message Service (SMS) 1. Introduction 2008 Beijing Olympics raise the aim of being a “digital Olympics”. In 2008, a great number of athletes, spectators and tourists all over the world will come to China, and that providing them with convenient mobile order service as well as query service can embody the feature of 2008 Beijing Olympics. Based on it, we try to develop this Olympics-oriented mobile game news ordering system. At present, the research on this system mainly aims at the items of table-tennis and badminton. Our final goal is to establish a quick and convenient game news information service system. The paper is structured as follows. In the next section the overall frame of the system is described. In section 3 we introduce machine learning and information extraction functions. Section 4 describes the designing and realization of the SMS service system. Finally, we will say a word about the prospects and promotion points of our game news ordering system. 2. The overall Frame of the system The system is mainly made up by two modules, the machine learning and information extraction module and the SMS service system module. Fig. 1. The overall frame of the system The task of the first module is to process the free Chinese game news texts to produce structured data as follows: With the machine learning we obtain the initial rules, after being evaluated and pruned, the rules will be put into the knowledge base of information extraction, and then the texts can be changed 459 into structured data through information extraction. The main task of the second module is to interact with the users and to receive and send SMS. It is made up by the server side and client side. 3. Machine learning and information extraction 3.1 Machine learning 3.1.1 Introduction of the program To generate rules, we used the machine learning method of from special to common and from bottom to top. Under our supervision, machine trained with the labeled corpus, then the initial rules are produced. More generalized rules are obtained through generalizing some items in the initial rules, and finally the exact and recapitulative rules are generated. During the first processing, the program distilled the labeled sentence from text. Then we used the technology of shallow parsing to delete unimportant words. Then we specially designed a …

Content truncated to 500 words