What People Tweeted About on Feb 22 2015

Design Visualization for Twitter Trends Data

Irene Ye Yuan
3 min readDec 4, 2018


What is the best way to visualize popular twitter trends in different U.S. cities?

With the raw data of popular Twitter topics across 60 cities across United State on February 22, 2015, designing an intuitive way to arrange and visualize the data so that users can explore and compare the data with respect to topics, locations, and time.

Explore the Data Set

  1. The locations of cities are not evenly distributed, which makes it hard to visualize the cities on a base map without confusing users.
  2. The distribution of the trend topics is sparse. After the top two popular trends, popular topics varied a lot across cities. The common popular topics are rare to see.
I quickly sketched up a concept for map based visualization.
To test the visualization, I built part of the visualization with the test data.
After playing with the test data, I realized the map based visualization doesn’t work, and switched to a grid view.

Final Design Overview

The final design is an interactive visualization with grid view of all the cities

Arranging by time zone

In this 10 by 6 grid, all the cities are first grouped into different time zones, and arranged from right to left, top to down, with the time zone order. The grid also provides a visual baseline for exploring similarities and differences between different cities.

Summary of twitter trends

Select Topic tab summarizes all the topics appeared in the data set, and arranged in the decreasing frequency order. By clicking one topic, users can pin it to the main visualization, which means that the topic will always be present in the visualization.

Exploring by hovering and playing

By hovering over a topic, the tooltip shows the topic description, while highlighting the same topic in other cities if exists. Users can also explore the data by time, either click the play button to play through all day, or drag the toggle to see a specific time.

Make it Alive in Processing

View Source Code



Irene Ye Yuan

HCI Researcher & Technologist, PhD Candidate @ GroupLens, University of Minnesota