The Rhythm of Food is a project that allows users to view Google search trends(within the last 15 years) about food in an interactive, visual way. Created by members of Google News Lab and Truth & Beauty, the goal of the project is facilitate analysis of food / recipe trends over time. The target audience of the project is the general public, as they want anyone to be able to make their own discoveries in food trends.
The link to the project: http://rhythm-of-food.net/
Sources:
The food related search terms come from FooDB (https://foodb.ca/) and the Google search data itself comes from Google Trends (https://trends.google.com/trends/?geo=US).
Processes:
The project takes data from Google Trends and simply displays them in a format that suits their goals. I am unfamiliar with the details of how the data within Google Trends is formatted, but I envision that there is some way to make queries to Google Trends to receive the number of searches for a certain keyword. Therefore, it does not seem that there was too much complex data processing / wrangling involved since much of the grunt work was already done by Google.
One aspect of the project that I could see being difficult is determining the subcategories that foods fall under. For instance, how can one know that egg nog falls under “Christmas foods”? The creators address this briefly, mentioning that they used Google Knowledge Graph for these issues.
Presentation:
The project is a single, long webpage that has interesting insights and visuals throughout. From the blog of Moritz Stefaner, one of the leaders of the design of the project, I found that the page is organized in a “Martini glass structure”, where the user is given an overview of the project, demonstrated a few key interesting features, and then given a dashboard to make their own discoveries. The creators made it a priority to not place the burden of finding interesting selections on the user. From this blog, I also found that this project was built using ES2015, webpack, react, Material UI, and d3 v4.
From the blog of Yuri Vishnevsky, I found some of the rationale behind the spiral structure of the visualizations. First, the smooth spirals really suit the theme of seasonality and change, as one can easily see how these trends change ebb and flow as the seasons repeat. Additionally, the spirals are designed to allow human brains to spot patterns much more easily compared to simple line charts. Combined with the fact that they used the universal calendar layout in these spirals– each week is a row, each year is a circular revolution– these visuals are easy for people to grasp quickly.
Question:
One question that I have is why they limited this project to only food. Although it does make sense given the way that food patterns change over time, I’m sure that there are many search trends for other categories that could be very interesting to analyze. Of course, I could be completely wrong and not know that they have already created such projects.
