Mapping ADS Usage

Skills used:
Python, D3, DataViz

1. Visualizing institutional representation at the 225th American Astronomical Meeting

How I Made the Map

Below is a map that visualizes the representations of different institutions at the 225th American Astronomical Society Meeting. I made it by using the Astrophysics Data System API to gather affiliation information for all authors who submitted abstracts to the conference. I then used Google's Geocoding API to find the location of the institutions. The JavaScript visualization library D3 helped me make a quick map to show the relative contributions of different institutions. If you mouse over a circle, you can see all members who authored an abstract. When we showed this at our booth, we hit the "loop" button to highlight a random institution every 5 seconds.

AAS Abstract Map

Click on a circle to view abstracts from people from that institution in ADS. A quick caveat: these locations were automatically geocoded using imperfect affiliation data and so some of them (hopefully not many) contain errors.

Hover over a circle to view the institution's name and a list of affiliated authors
Hover over a circle to view the institution's name and a list of affiliated authors

2. Global ADS Usage Visualization Using Google Analytics Data

This first gif illustrates worldwide use of the Astrophysics Data system measured in individual user sessions. It illustrates the working days of astronomers across the world.

While building the first visualization, I realized that I also had access to search data using the Google Analytics API. The bottom gif is my attempt to give an impression of the different types of searches made on the ADS during the same 24-hour period. It is a bit difficult to make out the text of the searches, and if I were going to improve this visualization, I would make it interactive and zoomable so that a user could choose a smaller geographic region before seeing the stream of searches made during a particular time period.