| With the arrival of the era of big data, data plays a more and more important role in social life. With the popularization of the Information System, in which the data construction as an important system engineering, we encounter many new problems in the construction process, such as distributed data resources, the lack of high value information etc. The army data construction is facing these challenges, at the same time there exists the problem of low efficiency of data utilization. How to dig out the troops, equipment and other knowledge from a large amount of historical military information data, so as to help troops to solve problems and provide decision support for the leader, is a significant work.With the rapid development of computer technology in recent years, data mining technology with the clustering analysis as the representative has become a very active research direction, where new theory and new algorithm have been proposed a lot. The combination of data field in spatial data mining theory and clustering algorithm which achieved good clustering effect, have broad application prospects. Because the shape of the data set after embedding the space is not standard, the scope of application of density clustering algorithm which can partition the irregular shape of the clusters is more widely.Therefore, aiming at the practical problems encountered in the participation of military data in engineering construction, combined with data mining method, the research mainly include the following aspects:(1) We put forward Macro Data Field and Data Situation theory. We introduce the basic concepts and content of the two in this paper, and clear the conceptual model and the application field of Macro Data Field, then ensure the main content and research methods of Data Situation;(2) We study the clustering algorithm by density peaks combined with the data field theory, by using the potential function to build the data field containing the whole information, and choosing the point having the extreme potential value as the initial point, then calculating the parameters of each cluster, finally catching the clustering results;(3)We use our military information resources data as the data source to analyze the situation, and mainly do the following work:(1)We deal with the data source by filling the missing data, smoothing noise and outlier removing data.(2)We build the model of situation analysis of the personnel allocation and variation, and use the density clustering algorithm to do the clustering analysis, which showed that the officers, the students of cadres, the students of sergeancies, and the employments fulfill the headcount of stuff easily, in contrast with the soldier and the sergeancy have missing stuff a lot.(3)We establish the model of the situation analysis of equipment collocation, and build the function of excess rate and total excess rate of the equipment, then define the hierarchical function to classify the height. The results showed that the military equipment collocation has regional difference; height after classification has a significant negative correlation with equipment integrated excess rate; army equipment full rate have no significant correlation with army personnel.(4) We establish the situation analysis model of equipment status, and definite the degree and rate of acquired equipment, then take a certain type of equipment as example to analysis, finally the results showed that the existence of regional differences of equipment status, as well as the correlation with height. |
