The Study And Construction Of Forest Ecological Station Timing Data Pattern Mining System

Forest ecological stations had accumulated much more data, and the data had been more complex. With the rapid development, data mining technologies can be used in the timing data of forest ecological stations to mine some valuable patterns hidden in the data. That has been drawn more and more attention of experts and manager in ecological field. To improve the management level of forest ecological stations, it is very important to apply the patterns, which include one-dimensional patterns, multi-dimensional patterns, single-factor patterns and multi-factor patterns to the actual forest ecological stations’production.Firstly, this paper describes the research status and significance of the Forest Ecological Station data mining, and then the related knowledge and mining algorithms of the pattern mining system are described in detail, such as FFT to find the cycle of time series, association mining algorithms to find the relation of different variables. Secondly, based on the above theories, the design of the forest ecological stations timing data mining system is put forward. After the implementation of the system, the system can mine some valuable patterns, such as cycle of time series, association rules and the clustering results. In the end, this paper analyses and evaluates the mined patterns. The innovation point of this paper is to put forward a cut points searching algorithm, which can divide the variable intervals in order to transfer the original numerical data to transactional data. That can improve the results of association rules. At the end of this paper, the problems existing in the research are summarized, and subsequent research work is also put forward.