THIRUVANANTHAPURAM: In a bid to shift to predictive policing using artificial intelligence, the CCTNS team of the State Crime Records Bureau has ventured into developing tools that could help investigators predict crimes, identify crime hotspots and detect links between various crimes, which otherwise, could be missed by cops doing it the traditional way. The basic model is expected to be completed this year.
The wing’s objective is to harness the huge data lying in the police repository and crunch it using AI algorithms. The AI models can do analytics, find patterns and then offer predictions.
Suppose thefts happen in certain pockets of the state in rainy season during a particular time window, the AI tool can spot the pattern and give a prediction. This can be done by collating data such as geographical locations, weather pattern and location of criminals, helping cops take action beforehand.
The tool can also sift through years of data and map the zones where crimes regularly happen, thereby identifying crime hotspots. What makes this a useful method is that the cops can monitor those areas using surveillance tools, cameras or mobile patrol units.
The algorithm, again, can help find minute patterns in crimes and establish links between those acts. This could be useful in unmasking the working of organised criminal rackets. By leveraging the big data, the close-knit working of the drug rackets can be exposed.
“We can use the tool to establish links between the buyers, peddlers, sellers and financial backers. The entire ring can be mapped and identified,” said I G P Prakash, CCTNS nodal officer. “AI-enabled models will surely help the force take proactive measures to prevent crime before it happens. It can also aid in swift detection of cases as well.”
Anomaly detection in crimes is another aspect that the tool can help with, especially while probing dead-end cases. The AI algorithm can also enable crime trend prediction on time-series models.