An Agent-Based Study On Customized Technology Of Personalized Travel Information

Now day, travel information service is usually studied from improving the quality of information and analyzing the influence of the information service. The integrity, accuracy and timeliness of information are considered. And the traveler's reflection of information is also studied. While traveler faces massive information, he can't analyze and handle information because of his limited sense. In this situation, it will appear extreme reflection, gathering reflection and so on. In this paper, the problem is considered from the angle of travelers'personal demand difference, and decision tree method is adopted to obtain traveler's personal request of travel information. Agent technology is also used to assist travelers handling information. The advanced information service is realized in this method. In addition, mobile computing method is used to construct the travel information service network.The content of personalized travel information service is analyzed in this paper such as the expression manner of information, demand, classification and strategy. The new content of personalized travel information is endowed from the angle of information service level. Using mobile equipment is proved to be the best manner to supply information service by comparing with others. Based on these researches, the architecture model of ITS information service network is established. And Agent technology is used to establish the model of information service terminal model. The function of the components is also defined.C5.0 algorithm of decision tree in machine learning is used to solve the problem of information demand. Information gain rate is used in C5.0 algorithm to divide records. It enhances efficiency and handling ability for large-scale database. Most importantly, the rules can be obtained. These rules are precisely the embodiment of traveler's demand of personalized information. The validity of this algorithm has been proven through an example.To analyze the effect of personalized information service to the dynamic network, Starlogo software is used to simulate the situation in this paper. By simulating the network which only contains two type travelers, namely ordinary...