Study on loneliness with OECD data

Inspiration:

I would like to choose a specific topic from the OECD data. One measurement caught my eye: Feeling Lonely.

What it does:

This app tries to determine the percentage change in the above measurement. It also reviews the results by geography and demographics, e.g., age, sex, and education. Meanwhile, it also shows the correlation between loneliness and other measurements to reveal a possible reason for improvement.

How we built it:

This app was built entirely with Plotly Studio. The first version was built by providing the project's goal, investigating global loneliness, and refining the graphs, layout, and color through improved specifications. Changes to the code were minimized to take advantage of Plotly Studio's AI feature.

Challenges we ran into

Apart from the chart showing the score by country, which comes from the first version, the styling has been refined. All charts were recreated. The AI tends to pick simple charts to create visuals, and the choice of parameters is suboptimal. It seems that the AI feature cannot suggest colors based on the topic or feeling. Meanwhile, it is difficult to revert to the previous version after making a change because there is no prompt history, so it is unclear if the same result will be returned. The global filters are sometimes chaotic.

Accomplishments we're proud of

Not a single line of code was written to build this. We just made some changes to the wording of the title and labels.

What we learned:

• Specify all the column names to be used to reduce iteration.
• Optimize the prompt with support from other AI tools, e.g., ChatGPT and Gemini.
• Specify a filter for the data used in each chart.

What's next for Study on loneliness with OECD data

No further planning yet.

Built With

• ploty