The Research Of The Bayonet Information Mining Technology

With the increase of the number of vehicle, the bayonet information of vehicle had a rapid increase. Through the data mining of the bayonet information, we can get a lot of neglected potential information, for example, find the congested road, the accompany vehicle etc., so as to achieve the effect of the intelligent traffic management. However, with the increase of bayonet information, there are more difficult in the storage, management, data mining of bayonet information. Therefore, mount bayonet information storage and processing has important application significance.Firstly, this paper first introduces the development of intelligent transportation, data mining process and the technology of big data. In order to completing the goal of intelligent traffic management through the data mining of bayonet information, in view of the problems of the original IOE framework storage pressure, high cost of equipment, real-time processing delay, this paper decided using HBase database to storage bayonet information and offline data mining, using Flume-kafka completed collection and Storm to complete the real-time processing. For data storage and offline data mining, this paper aim at accompany car. Through the determination of the definition and mining task of accompany car as well as the design of HBase table, complete the storage of bayonet information and find the accompany car quickly. For real-time process, this paper aim at the set of sign vehicle. Design profit algorithm after process real-time bayonet information. Lastly, the result indicate that compared with the traditional IOE framework, the efficiency of retrieval accompany car has promote 260 times, and the result of retrieval is more comprehensive. At the same time, in consider of the unreasonable of space-time and the error rate of the recognition system, complete the process of real-time bayonet information.