The power of public sentiment in city building
A doughnut chart displaying public sentiment is both a fun and powerful way to show a snapshot of how citizens feel
Displaying public participation in real-time creates transparency in citizen engagement and city building
What do we do with your feedback?
At Milieu, we incorporate your feedback to create insightful analytics about citizen engagement in city building. The doughnut chart is one good example of how we display what citizens are feeling and saying, and publicly share the real-time results of participation.
Formally, the method is called sentiment analysis. The goal is to computationally determine people’s sentiments, opinions, or feelings towards a certain topic. Many companies will collect this information for various purposes, some good and some more nefarious. For us, it’s about bringing together the voices of citizens, and making that collective voice visible to decision makers and the public. Understanding how people feel about their city can be an important insight as to what needs to change.
Let’s take a closer look at the doughnut chart
In our analysis we transform a user comment into five different numbers. We represent an individual’s Joy, Anger, Fear, Sadness, and Disgust for each comment that they leave on a scale of zero to one. For example, for the 190 Richmond Rd. development we calculated the total emotional output from the comments collected.
This is what the actual chart looks like. Each slice of the doughnut represents that emotion’s share of the total emotional content. For example, at 190 Richmond Rd, 31% of the total emotional content was Joy (teal).
From here we can start asking questions like, what was talked about in the comments that had the most Joy in them? We might learn things like: people are happy about elements of the design, or the fact that affordable housing is mentioned in the site application. We can gain insight into the things that address the key concerns residents might have.
In contrast, we can learn what people would like to change by looking at comments with anger, disgust, or fear. We might learn that people are worried about the height of the building, or that they would like it to have some set-back included.
However, none of this is news. Even before fancy cognitive algorithms existed, it wouldn’t be hard to tell if people were happy about affordable housing, or upset about the height of a building. What is new, is that with cognitive computing we can determine these things using a consistent and less biased process, where results are made available in real-time to the public and decision makers.
