Abstract: Crime rate is increasing very fast in India because of increase in poverty and unemployment. With the existing crime investigation techniques, officers have to spend a lot of time as well as man power to identify suspects and criminals. However crime investigation process needs to be faster and efficient. As large amount of information is collected during crime investigation, data mining is an approach which can be useful in this perspective. Data mining is a process that extracts useful information from large amount of crime data so that possible suspects of the crime can be identified efficiently. Numbers of data mining techniques are available. Use of particular data mining technique has greater influence on the results obtained. So the performance of three data mining techniques { J48, Nave Bayes and JRip will be compared against sample crime and criminal database and best performing algorithm will be used against sample crime and criminal database to identify possible suspects of the crime. Data mining is a process of extracting knowledge from huge amount of data stored in databases, data warehouses and data repositories. Clustering is the process of combining data objects into groups. The data objects within the group are very similar and very dissimilar as well when compared to objects of other groups.

 

Keywords: Criminal database, Crime investigation, Data Mining, J48, JRip, and Na´ve Bayes.