Abstract:

For further solving the problem of normative emergency plans,the paper combines the text mining technique with the emergency plan preparation to propose an improved DBSCAN algorithm.The algorithm provides intelligence reference groups for government staffs.In process of emergency plan preparation,the Least Square Fit is taken to fit text similarity curve.During the clustering iterative process,the adjustable neighborhood threshold Eps was calculated so as to improve the precision and the recall percentage.The simulation results show that the presented method is efficient.

Key words: DBSCAN algorithm;text similarity;text management;plan preparation