Data Mining Method in Forest Engineering Based on Big Data and Internet of Things Technology

Abstract

A large amount of forest data in forestry engineering leads to that data management is a very
complex work. Data mining mainly includes four parts: clustering analysis, prediction
modeling, association analysis and anomaly detection. In order to get the potential valuable
information from the complex data, it is necessary to deeply study and flexibly use the data
mining algorithm. Based on spark distributed framework, this paper focuses on the maximum
frequent itemset mining algorithm and density clustering mining algorithm. In the aspect of
frequent itemset mining, because of its high value, advanced information is hidden in long
frequent items. Therefore, mining maximal frequent itemsets has higher value. After combining
the advantages of the existing algorithms, a recursive deep path search is proposed to generate
the maximum frequent item candidate set at one time. Then, the candidate frequent itemsets are
sorted by length first. Then the improved process of superset test is recycled. The experimental
results show that the improved algorithm optimizes the pruning and dimensionality reduction of
the data set, and reduces the scale and mining times of the candidate item set. This method
solves the problem of low efficiency of the existing maximum frequent mining algorithm when
the amount of data is large and the dimension is high.