Each year, this municipality receives approximately 400,000 public space issue reports (known as Meldingen Openbare Ruimte or MORs) from citizens. These reports cover a variety of urban challenges, including waste management issues, disturbances caused by homeless individuals, and damaged infrastructure. The primary challenge in managing these reports has been the manual and labor-intensive process of analysing feedback on how the municipality handled each case. This conventional approach was time-consuming, often leading to an incomplete assessment due to the vast volume of unstructured text data.
Strategic challenge
The strategic challenge lay in transitioning from this manual, resource-intensive process to an automated, efficient system. The goal was to employ advanced AI techniques, particularly large language models (LLMs), to allow for automatic categorisation of responses to MORs.
Approach
The project involved several key steps:
1. Leveraging AI to automatically analyse and classify MOR responses
Large language models were leveraged to automatically convert the unstructured text describing the complaint resolution into categories such as "Immediate Action Taken," "No Enforcement Available," and "External Agency Referral", “Extra surveillance during the coming period”, “Report not processed due to malfunction in reporting system”, etc. This way, structured data is obtained that can be used for further analysis.
2. Generation of statistics on complaint handling
The classification results were converted into detailed statistics. Key performance indicators, including the frequency of each complaint resolution method, response times, and satisfaction ratings, were automatically compiled for comprehensive analysis.
3. Automated generation of PowerPoint reports for delivering final insights
-
After completing the statistical analysis, an automated system generates a comprehensive PowerPoint presentation encapsulating all crucial insights.
-
This automation streamlines the process of report generation, transforming data into a visually appealing and easy-to-understand format.
-
The PowerPoint presentation includes graphs, charts, and summaries that provide a clear narrative of the data, making it easier for teams and stakeholders to review and make informed decisions. This saves significant time and effort in report preparation.
Results
Integrating AI and LLMs in the MOR analysis process led to significant advancements:
-
Increased efficiency: Automation dramatically reduced the time and labor involved in categorising MOR responses, enabling quicker and more thorough analysis.
-
Enhanced insights: The municipality gained clear insights into its response types and effectiveness in handling public space issues.
-
Informed decision-making: This new, data-driven approach facilitated more strategic and informed decision-making, improving public services and enhancing citizen satisfaction.
Conclusion
The innovative integration of AI and large language models has notably advanced the analysis of public space complaint handling in this major European city. This project sets a new standard for municipalities and public service sectors, showcasing the potential of AI in automatic text processing and reporting.
