Keeping tabs on thousands of crimes happening every day can be tough. But we're not doing a great job of it. In fact, crime rates are climbing because we don't pay enough attention or collect enough data. This is a problem because people need to know where it's safe to go, police need to catch criminals, and tourists need to plan their trips.

The National Institute of Justice does provide crime data for the whole country, but it doesn't give us details down to the local level, like cities or districts. So, we've found a way to fill that gap. Our method uses online Bangla news sites to find crime data, figure out the crime ranking of different areas, and even predict where crimes might happen in the future.

Here's how it works: We gather crime news from various online Bangla newspapers, then use a special process to find keywords and locations from the articles. We use a technique called cosine similarity to match similar crimes. Then, we use a tool called a Naive Bayes classifier to sort crimes into categories. Finally, we use a custom crime prediction model to guess where the next crime might happen.

The goal is to give people a heads-up about the safety of a specific location. By doing this, we hope to help people make smarter decisions about where to go and when.

Predicting Safety: How Online News Can Help Us Avoid Crime Hotspots

Actions