Design And Implementation Of Forest Ecological Station Data Mining System

Along with the develop of ecological construction, our country has established many Forest ecosystem research stations. The role of stations is long-term ecosystem locating-observation, including solar radiation, air temperature, soil temperature, precipitation, evaporation, wind speed and so on. Because of the multiple observation indicators, the huge amount of data over time and the various spatial distribution of station, how to extract useful information from the data is one of the urgent problem needed to address. With the advent of the era of big data, the Internet, databases and data mining technology is changing rapidly. More and more researchers have begun to focus on how these advanced technologies can be applied in information extraction from forest ecosystem research station data. With in-depth analysis of the data, we could find not only the internal development and succession mechanism of ecological systems, but also the dynamic balance of forest ecosystems. These findings will help us understand how forest impact other factors in biogeochemical cycles.First, this paper elaborates the research status and practical significance of data mining technique in the field of forest ecosystem research station data. Second, the paper describes in detail the forest ecosystem research station data mining system and the algorithms used, including the use of Fourier function to calculate the index cycle, with the association rule mining index relationship between the clustering algorithm using excavated clustering center. Third, the results of system application in experiment area are compared and evaluated. Finally, future study direction in the forest ecosystem research station data mining is analyzed based on current research progress. The difficulty lies in the realization of several related algorithms such as clustering algorithm K-Means, the improved algorithm X-Means, GA-based discrete data and then use Apriori algorithm for mining association rules, as well as improved algorithm FP-Growth and the like. The innovation is that the two improved algorithm in this field use.After the system test mining results obtained with a certain reliability and relevance, in line with the actual requirements of forest ecology station.