Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

---

Introduction: In recent years, technology advancement and the growth of information technology in organizations have provided a huge source of data stored in the field of drug-related offenses. Analyzing these data and discovering hidden patterns in it can help detect and prevent the occurrence of crimes in this area. This paper aimed to identify the susceptible people to drug trafficking in Sistan and Baluchestan province and discover patterns of crime using data mining techniques.
 Method: The present study was conducted on data of 467 drug offenders in Sistan and Baluchestan province who have committed drug trafficking in the years 2011 to 2020 by available sampling. CRISP-DM methodology was used to build a prediction model. Also, Support Vector Machine (SVM), Naïve Bayes, Logistic Regression nad Decision Trees have been used to predict people at risk and Apriori Algorithm has been used to extract crime patterns.
Findings: The pattern mining algorithm extracted over 20 crime patterns with a precision of over 80%. The results of the evaluations show that the IBK classifier can accurately identify 84 % of the people at risk.
Discussion: A system for identifying people susceptible to drug trafficking can be designed using the model made by the IBK classifier. In addition, the results of the predictions by the above mentioned system as well as the extracted hidden patterns can help police, judiciary and social workers to identify people at risk and reduce drug-related crimes.

Keywords: Drug offenses, Data mining, Support vector machine (SVM), Crime patterns, Crime prediction

Full-Text [PDF 607 kb]   (1099 Downloads)    

---