Temporal Pattern Mining in Illicit Drug Seizures: A Comparative Analysis of Heroin and Methamphetamine
Article Sidebar
Main Article Content
Abstract
The increasing number and changing trends of heroin and methamphetamine in Sri Lanka are a major problem for the health, law enforcement, and policy-making of the community. These illicit substances have been displaying significant growth and changes through the weight categories over the period between 2015 and 2024, indicating the necessity to intervene and solve the problem in a timely and data-driven manner. The analysis of data on seizures in this study was based on anomaly detection, change-point analysis, modeling of weight-category transition, and yearly clustering to reveal structural changes, trends, and patterns of interaction in drug-related activities. The cases of heroin increased steadily between the year 2015, when the number of cases per month was around 120 cases, and 2024, when the number of cases exceeded 450 cases per month, and the number of cases of methamphetamine increased after 2022 and had over 300 cases per month in 2024. The anomaly identification in June-September 2020 and January 2024 showed that the deviations were substantial, and this meant that the heroin activity was largely interrupted. The change-point analysis revealed the changes in the trends of heroin in 2018-2019 and 2025, but not the changes in methamphetamine, which began to change in 2022. Heroin incident at transitions between weight categories decreased (< 2 g) and increased (2-5 g and 5-10 g), indicating a trend of increasing typical seizure sizes. The annual clustering indicated that 2020 and 2022-2024 will be years of high activity of heroin, and 2024 will be a year of high activity of methamphetamine. This study will help offer a scalable decision-support model to policymakers, forensic labs, and law enforcement agencies by offering an integrated approach to drug activity change monitoring, strategic planning, and resource allocation. The framework also creates a base of predictive systems in the future that will take into account socio-economic, demographic, and real-time surveillance information to improve proactive intervention measures.
Article Details
Copyright (c) 2026 Bakmeedeniya AHMTC, et al.
This work is licensed under a Creative Commons Attribution 4.0 International License.
1. Palipane J. Sri Lanka has arrested tens of thousands in drug raids, criticized by the UN human rights body. AP News. 2024 Jan. Available from: https://apnews.com/article/sri-lanka-drug-raids-human-rights-4a2c267c246e89fe919335d927f68887
2. National Dangerous Drugs Control Board (NDDCB), Sri Lanka. Annual report 2023. Sri Lanka: NDDCB; 2023. Available from: https://www.nddcb.gov.lk/annual-report.php
3. Darshana N, Wijesinghe CJ, Silva PVD. Assessment of characteristics and exposure to vulnerable factors for drug use among male illicit drug users in Sri Lanka: a multicenter cross-sectional study. Indian J Community Med. 2022;47(2):285-291. Available from: https://doi.org/10.4103/ijcm.ijcm_712_21 DOI: https://doi.org/10.4103/ijcm.ijcm_712_21
4. Jayamaha AR, Dharmarathna ND, Herath NDM, Ranadeva NDK, Fernando MM, Samarasinghe KL, et al. The pattern of substance use and characteristics of the individuals enrolled in residential treatment at selected rehabilitation centers in Sri Lanka: a descriptive cross-sectional study. Subst Abuse. 2022. Available from: https://doi.org/10.1177/11782218221100823 DOI: https://doi.org/10.1177/11782218221100823
5. Harmon J. Forensic laboratories as drug policy partners. Police Chief. 2024 Jan. Available from: https://www.policechiefmagazine.org/forensic-labs-drug-policy-partners/
6. Degenhardt L, Whiteford HA, Ferrari AJ, Baxter AJ, Charlson FJ, Hall WD, et al. Global burden of disease attributable to illicit drug use and dependence: findings from the Global Burden of Disease Study 2010. Lancet. 2013;382(9904):1564-1574. Available from: https://doi.org/10.1016/s0140-6736(13)61530-5 DOI: https://doi.org/10.1016/S0140-6736(13)61530-5
7. United Nations Office on Drugs and Crime. World drug report 2024. United Nations; 2024. Available from: https://www.unodc.org/unodc/en/data-and-analysis/world-drug-report-2024.html
8. World Health Organization. WHO Expert Committee on Drug Dependence: forty-first report. Geneva: World Health Organization; 2019. (WHO Technical Report Series, No. 1018). Available from: https://iris.who.int/server/api/core/bitstreams/d1e75e95-0021-4edd-85da-e5a5c923dd29/content
9. Hewamalage H, Ackermann K, Bergmeir C. Forecast evaluation for data scientists: common pitfalls and best practices. Data Min Knowl Discov. 2023;37(2):788-832. Available from: https://link.springer.com/article/10.1007/s10618-022-00894-5 DOI: https://doi.org/10.1007/s10618-022-00894-5
10. Bousqaoui H, Slimani I, Achchab S. Comparative analysis of short-term demand predicting models using ARIMA and deep learning. Int J Electr Comput Eng. 2021;11(4):3319. Available from: https://ijece.iaescore.com/index.php/IJECE/article/view/23508 DOI: https://doi.org/10.11591/ijece.v11i4.pp3319-3328
11. Li G, Yang N. A hybrid SARIMA-LSTM model for air temperature forecasting. Adv Theory Simul. 2023;6(2). Available from: https://doi.org/10.1002/adts.202200502 DOI: https://doi.org/10.1002/adts.202200502
12. Gers FA, Schmidhuber J, Cummins F. Learning to forget: continual prediction with LSTM. Neural Comput. 2000;12(10):2451-2471. Available from: https://doi.org/10.1162/089976600300015015 DOI: https://doi.org/10.1162/089976600300015015
13. Song X, Deng L, Wang H, Zhang Y, He Y, Cao W. Deep learning-based time series forecasting. Artif Intell Rev. 2024;58(1):23. Available from: https://link.springer.com/article/10.1007/s10462-024-10989-8 DOI: https://doi.org/10.1007/s10462-024-10989-8
14. Butt UM, Letchmunan S, Hassan FH, Koh TW. A hybrid of deep learning and exponential smoothing for enhancing crime forecasting accuracy. PLoS One. 2022;17(9):e0274172. Available from: https://doi.org/10.1371/journal.pone.0274172 DOI: https://doi.org/10.1371/journal.pone.0274172