top of page

Analysis of Incident Narratives and Temporal Patterns from Thunder Bay Police Service Reports

Unveiling Patterns in Thunder Bay Police Incidents: A Comprehensive Analysis of Incident Narratives and Temporal Trends (2021-2023 Q1)


Introduction:


This comprehensive research report presents an analysis of the incident narratives and temporal patterns documented in the Thunder Bay Police Service reports from the second quarter of 2021 through the first quarter of 2023.


Methodology:


The data was provided in quarterly CSV files for the years 2021, 2022, and the first quarter of 2023. Each file contained an 'incidentType' column, a 'narrative' column, and a 'date' column. The 'narrative' and 'date' columns were analyzed to identify common themes and patterns within the incidents.


Analysis of Incident Narratives:


The narrative column was analyzed using a natural language processing technique called n-gram analysis, specifically focusing on bi-grams (two-word combinations) and tri-grams (three-word combinations). This helped us identify frequently occurring phrases in the incident narratives.


Findings:


1. Motor Vehicle Collisions:


'Motor vehicle collision' was a frequently mentioned narrative throughout all periods.

  • Associated incident types often involved collisions with no injuries or any type of motor vehicle collision.

  • This suggests a consistent pattern of traffic incidents requiring police response.

2. Mental Health and Police Assistance:

  • 'Mental health act' and 'police assistance' were also frequently mentioned narratives.

  • Associated incident types often involved welfare checks, assistance to other agencies, or the public.

  • This indicates a substantial number of incidents related to mental health issues and general police assistance.

3. Unwanted Person:

  • The 'unwanted person' narrative often involved incidents where there was a potential for violence or no threat of violence.

  • This suggests a regular occurrence of incidents involving individuals who are perceived as unwanted in certain situations.

4. Domestic Dispute:

  • 'Domestic dispute' narratives were associated with incidents involving weapons or disputes where the suspect was not present.

  • This indicates a significant number of domestic incidents requiring police intervention.

5. Follow Up:

  • The 'follow up' narrative was commonly associated with incidents requiring checks for follow-up activities.

  • This could include ongoing investigations or situations requiring further police attention.

Analysis of Temporal Patterns:


The 'date' column, which includes date and time information, was analyzed to identify temporal patterns in the incidents. A heatmap was generated to visualize the number of incidents by hour of the day and day of the week.

Key findings include:

  • The heatmap showed that the number of incidents varies by hour and day, which could reveal patterns in when incidents are most likely to occur from 2021 Q2 to 2023 Q1 or from April 2021 to the end of March 2023.

  • Darker colours on the heatmap, which indicate higher numbers of incidents, can help identify specific times when incidents are more frequent. For instance, late-night and early-morning hours (around 0-5) on Sunday (represented as 6 on the heatmap) had a relatively high number of incidents, indicating that these times might be particularly challenging. In contrast, midday hours (around 10-14) during the weekdays showed a relatively lower number of incidents.

Conclusion:


This analysis provides valuable insights into the types of incidents that the Thunder Bay Police Service frequently responds to and when they are most likely to occur. These insights can help inform strategic planning and resource allocation, contributing to the continued safety and well-being of the Thunder Bay community. The recurring themes of motor vehicle collisions, mental health-related incidents, police assistance, unwanted person incidents, and domestic disputes underscore the complex and diverse nature of police work in Thunder Bay. Moreover, the temporal patterns provide additional context and could be useful in planning shifts and resources.

Comments


bottom of page