Full Day Course: Event Analytics and the Visualization of Temporal Event Sequences

Catherine Plaisant, HCIL, University of Maryland

June 30th 2017, Telecom ParisTech, Amphitheatre B310, 46 rue Barrault, Metro Corvisart.
See https://www.telecom-paristech.fr/telecom-paristech/adresses-acces-contacts.html

Registration mandatory. To register, click here by June 20.

Updated and dynamic page: https://go.umd.edu/plaisant-inria

Sequences of events are part of people's life, their travel, hospital visits, or web browsing experiences. Exploring collections of event sequences can be challenging for computer professionals and non-experts alike. In this full-day course, I will review visual approaches developed to handle temporal event data, with demonstrations and case study reports, and will introduce some of the tools we developed at the University of Maryland to make event sequence analysis more accessible to all. We will look at prototypes that might serve as inspiration for new research, and dive in more details on a few tools that may be useful for Paris-Saclay students and researchers to analyze their own data, such as EventFlow and Coco for Cohort Comparison.

Register for the course here. You will be invited to download the demonstration software, and to bring samples of your own event-sequence data to analyze. For PhD students from Paris-Saclay, the course can be validated as one day of "formation scientifique" by the School of Computer Science.


  1. Bring a laptop
    • EventFlow is in Java so easy to install, make sure you have a recent version of Java installed.
    • Coco requires python packages: flask (http://flask.pocoo.org/) and scipy (http://www.scipy.org/). You’ll get to play with Coco only the end of the day… so download and Install Python in advance if you want to speed up the general install process.
    • You will be encouraged to work in pairs so I expect that you will help each others to install the 2 prototypes we will use
  2. Bring earphones if possible (useful to listen to video demos on your own during exercises)
  3. IMPORTANT Sign this form which will allow you to use the software for the course (print or mail it to me). it is a long form, but mostly says you will not use the tools for commercial purpose, and won't disseminate it…

Once I get your form you will get the passwords to use the tools :)

  1. Think of possible data you might want to look at and questions you may have about this data
  2. Optional: Create or bring a sample (keep it small for that day!).

The data format is very simple. It is a 5-column tab-delimited text file .


You will need a identity card or passport to enter Telecom ParisTech due to security issues.

Wifi will be provided either with Eduroam or a guest access.

Lunch will take place at the local restaurant. You can pay cash (about 10-15€) or - and this is new and updated - with a Credit Card too.

Les personnels et étudiants de Paris-Saclay peuvent avoir une reduction a condition de montrer leur carte.

Short Bio

Dr. Plaisant is an internationally respected research leader in human-computer interaction and information visualization. She is the Associate Director of Research of the Human-Computer Interaction Lab (HCIL) of the University of Maryland. She co-authored with Ben Shneiderman the 4th, 5th and 6th editions of Designing the User Interface, a major textbook on the topic of Human-Computer Interaction. Her accomplishments span the interface development lifecycle, with major work in requirements gathering, interface design, and evaluation. In 2015 Dr. Plaisant was elected a fellow of the ACM CHI Academy.


10:00 Introduction to Event Analytics and Visualization of Temporal Event Sequences
10:30 Demonstrations
12:30 Lunch break
14:00 Lab session on EventFlow
15:15 Break
15:30 Lab session on Coco
16:45 Wrap-Up and Conclusion
17:30 End


You can email questions/bug reports etc. to eventflow.umd@gmail.com, and we hope you will send us a summary of your findings, wishes and lessons learned.


Please include the CoCo URL hcil.umd.edu/coco and references to our papers in your own papers and presentations if CoCo was used to analyze data or screens are shown

For more information, contact Jean-Daniel Fekete.