Detecting Approximately Duplicate Records in Database

摘要

The existing database system data quantity is huge, many of which are repeated data. Using the traditional approach for detecting approximately duplicate records to find similar duplicate records in the database will involve very large time complexity and space complexity, unable to obtain very good results. This chapter presents a method based on improved genetic neural network approach for detecting approximately duplicate records, using genetic algorithm to optimize the network's initial weights; and then using the BP algorithm to train the detection data to obtain network model. The experimental results show that this method can effectively solve the huge amount of approximately duplicate record data detection problem.