Hate Crime Analysis: A Comprehensive Study Using Terno, An AI- Powered Data Scientist
The Double-Edged Impact of Migration
Over the past decade, the world has witnessed a sharp rise in global migration. People are moving across borders in search of safer environments, better opportunities, and brighter futures.
While this blending of diverse cultures and communities enriches societies, it has also given rise to an unsettling reality—an increase in hate crimes.
Hate crimes do not stem from a single issue or target just one community. They cut across gender, sexual orientation, race, religion, ethnicity, language, and more, leaving lasting scars on both individuals and societies.
In this study, we set out to uncover the deeper causes of hate crimes in Toronto, Canada, and, more importantly, explore meaningful ways to combat them. Check out the full analysis here.
Dataset
Toronto Police Service Public Safety Data Portal
Table of Contents
- Dataset
- Hate Crimes: A Growing Concern
- Terno AI
- Methodology
- Findings by Terno AI Using Visuals
- Insights by Terno AI
- Recommendations by Terno AI
- Conclusion
Hate Crimes: A Growing Concern
- Increased Migration
More people are moving across borders in search of safety and better opportunities. Some citizens see migrants as threats to jobs, culture, or security.
- Rise of Woke Culture
Progressive movements have encouraged people to express their identities openly. This shift has empowered many but also triggered pushback from others.
- Backlash and Intolerance
As diversity gains more support, resistance has also grown. Intolerance toward people from different races, religions, genders, or sexualities has increased.
- Growth of Extremist Sentiments
Fear, misinformation, and cultural clashes are fueling hateful ideologies. These tensions often lead to hate crimes that damage both individuals and communities.
Terno AI
- Terno AI is a super-fast AI-backed data scientist that can help you generate data-driven insights and recommendations.
- It’s accurate, reliable, and helps you save valuable time and resources.
- This is an amazing tool for any organization that looks to make data-driven decisions for its business.
Methodology
- Download the dataset from the Toronto Police Service Public Safety Data Portal
- Connect to Dataset - We can connect our dataset to Terno AI either by uploading the dataset to Terno AI or attaching it along with the prompt.
- Here we have attached it along with the prompt.
Prompt: Use the attached dataset. Give insights into which are the most prominent factors for hate crimes. There is time stamps for events, there is also geographic information. I want to know if time and geography also a factor in these crimes. Show trends and numbers across various years and months using charts. Use donut chart to show how many arrest were made and use pie chart to show percentage of bias reasons. I need to build a dashboard which includes all of this, and also neighborhood wise numbers, changes in numbers year to year . Create charts and other visualizations you feel are important apart from this, to present our dashboard effectively. All visualizations should be neatly formatted, make changes to column names to make it easy to understand. Do not use column name as such in visualizations, make good headings for all the charts. The formatting should be perfect with no overlapping text or cut in the corners.
Findings by Terno AI Using Visuals
Bias Factors
- Religion-based incidents (43%) and race-based incidents (27%) dominate, with religion slightly ahead.
- Sexual-orientation bias (≈13%) is the next most common, followed by ethnicity, gender, language, and disability.
Temporal Trends
- Overall incident counts remained relatively stable between 2014 and 2018, but began rising steadily from 2018 onward, indicating a recent upward trend in hate crimes.
- Monthly seasonality reveals a clear pattern, with incidents peaking during the summer months and reaching their lowest levels in winter.
Geography
- Certain neighbourhoods account for the highest incident counts.
- Police-division‐level bar charts confirm these local “hotspots” and highlight several divisions with disproportionately high numbers of incidents.
Arrests
About 22% of all recorded hate-crime events result in an arrest, as shown in the donut chart.
Insights by Terno AI
- Bias Motivations - Race and Religion biases are by far the most common, accounting for over half of incidents.
- Arrest Outcomes – Approximately 22% of hate‐crime incidents result in an arrest.
- Time Trends – Monthly incident counts show clear peaks in mid-year months.
- Neighborhood Hotspots – A small number of neighborhoods account for the majority of incidents.
Recommendations by Terno AI
Prompt: What are the recommendations and conclusion we can make from these insights
Terno Response:
To handle the rise in hate crimes in an effective manner, we need a multifaceted approach. Given below is a breakdown of the key steps that can help make communities feel safer and inclusive:
- Smarter Resource Allocation & Patrol Strategy: In the top 5 identified hot-spot neighbourhoods, we can give more focused law enforcement efforts. We can also establish mobile liaison units for flexible, community-focused policing in these areas.
- Strengthen Community Engagement & Prevention: To strengthen community engagement, we can conduct anti-bias education workshops by collaborating with faith-based organizations, cultural groups, and schools. Run seasonal public campaigns during times when the crimes are more prevalent — to increase awareness and reduce prejudice.
- Enhance Training & Internal Protocols: Develop training programs for officers in cultural sensitivity, de-escalation methods, and recognize early signs of bias. Build a standardized Hate-Crime Response Protocol for consistent investigation and victim support.
- Improve Data Transparency & Performance Tracking: Create a real-time dashboard to track hate crime data by neighbourhood and motivation. Set quarterly performance goals and track progress to ensure accountability and continuous improvement.
- Prioritize Victim Support & Outreach: We could set up a dedicated hate-crime victim liaison unit to provide compassionate, timely assistance for the victims. Partnering with community organizations can help raise awareness about a confidential reporting hotline, making it safer for victims to come forward.
Conclusions
- Using Terno AI was a ground-breaking step in the analysis. What would normally take hours to visualize was done in a matter of minutes — helping us save valuable time and effort.
- Moreover, the insights given by Terno AI were not only fast but also accurate and actionable. Terno AI helped us discover key patterns in the data that would have been tough to find out otherwise. This enabled us to make decisions that are well and truly data-driven.
- In a world where clarity matters, Terno AI makes analysing complex data feel easy. See our full analysis here.
