- Open administration
The Open Administration Awards 2024 will be presented on May 20 in El Prat de Llobregat: save the date
The AOC has been conducting satisfaction surveys among users of its services for over 9 years to collect their evaluation, detect the strengths of the services and areas for improvement. During this period we have collected more than 2 million surveys and 500.000 comments.
Only the 2024 237.742 surveys have been collected, with 51.573 qualitative comments that had to be analyzed, summarized and structured. For 5 months, we have incorporated generative artificial intelligence (AI) to streamline this task. Until now, it was done manually, and only a sampling of the comments because we did not have the resources to analyze them all. Approximately, it took us 1 minute for every 5 comments: this means that if we analyzed all the comments (4300 on average), we would have dedicated more than 14 hours per month.
Generative AI is a very useful tool for extracting knowledge
The use of generative AI represents a significant advance in efficiency: it allows us to analyze all comments on a monthly basis. At the same time, it generates executive summaries containing key knowledge (insights) to find out what brings the most value to users and what we need to correct or improve. In this way, it is an important tool for applying the strategy of people-centered services and the continuous improvement of AOC services.
The process and results of the implemented algorithm are as follows:
- Classification of feelings: AI can analyze comments and determine whether they are positive, negative, or neutral. This is done using natural language processing (NLP) techniques that identify key words and phrases associated with different sentiments.
- Topic extraction: AI can identify common themes in comments. For example, if many comments mention “access issues” or “quality of service,” AI can group those comments under those themes.
- Automatic summary: AI can generate summaries of comments, highlighting the most important points. This is especially useful when there is a large volume of comments.
- Data visualization: AI tools can create graphs and tables that show the distribution of positive and negative comments, as well as the most mentioned topics. This makes it easier to understand the data at a glance.
- Pattern detection: AI can detect patterns in comments that might not be obvious to the naked eye. For example, it can identify whether there is a correlation between negative comments and certain time periods or specific events.
These capabilities allow the AOC to better understand user opinions and make informed decisions to improve its products and services, in an agile and rapid manner.