A NOVEL KEYWORD SEARCH METHOD FOR ENVIRONMENTAL REPORTS BASED ON IMPROVED APRIORI ALGORITHIM

摘要

In recent years, more and more emergency en-vironmental problems have been widely discussed because of the frequent occurrence of emergency en-vironmental events. Due to the large amount of in-formation, it is difficult for users to obtain the con-tent of emergency environmental events. Therefore, this paper proposes a keyword retrieval method of emergency environmental events based on improved Apriori algorithm, in order to improve the retrieval efficiency of environmental events keywords. In this method, the report of environmental emergencies can be divided into four types:"pollutant discharge", "production safety accident", "traffic accident" and "natural disaster".Lingo clustering algorithm is used to extract the class tags related to the four types of environmental emergency reports, and all the docu-ments are assigned to the tags corresponding to the class tags to complete the clustering. The method uses Boolean matrix to improve Apriori algorithm, by mapping clustering results to form a Boolean ma-trix and setting the minimum support to obtain fre-quent itemsets, that is, emergency environmental in-cident reporting keywords. The experimental results show that the research method can effectively re-trieve the keywords of environmental emergencies reports. When the minimum support is 3, the re-trieval accuracy and recall rate are the highest, which improves the efficiency of data mining and realizes the effective acquisition of unexpected environmen-tal events.