Collective Craving logo

A Twitter Visualization of Human Cravings.

Collective Craving is designed to be viewed on a larger screen. Please resize your browser, or if you are in mobile please revisit this link using a laptop or desktop.

What is Collective Craving?

Collective Craving is an interactive Twitter visualization of human cravings during specific events. It has collected almost one million tweets containing the keyword ‘craving’, captured on 13 major US public events and holidays from Fall 2015 to Spring 2016.

The visualization provides users with a macro-level comparison that shows the top 50 most craved foods during those events and how the popularity of a given food changes over time. It also lets users be immersed in a micro-level exploration of every event, within a specific time constraint. At the core of this project lies an investigation into what kind of insight we can gain from the arbitrary, everyday data that we usually don’t care about, and how to turn it into a more meaningful knowledge. This project tries to capture the fleeting moment of human desire and cultivate the uniqueness of cravings as something ephemeral and mutually experienced. By creating and analyzing the database of human cravings, this project aims to evoke interest in the socio-digital era through visualizing our food culture as depicted by online users.

The Making of Collective Craving:

1. Data Collection

To create a database of human cravings, there were two approached tested: getting data from manual input (active observation) and from automatic input (passive observation). During the beginning phase of this project, the initial prototype was a physical installation to instigate and capture the user's inner food cravings. This was highly effective to attract more audience on a short exhibition setting because it involved real food display (pizza, burger, steak, sushi, and bacon). But at the same time, the on-demand dataset became very limited because of the small population of user, and it didn't accurately capture people's inner cravings in real time.

The second alternative was to do passive observation by scouring data from the internet. So I started scraping for tweets containing the keyword 'craving' one day before, during, and one day after these major US events and holidays from Fall 2015 to Spring 2016 (along with their total number of tweets):

  • Thanksgiving (24-27 Nov 2015) 73,459
  • Christmas (24-26 Dec 2015) 62,550
  • New Year’s (31 Dec 2015-2 Jan 2016) 72,584
  • Super Bowl (6-7 Feb 2016) 59,903
  • Lunar New Year (8 Feb 2016) 30,488
  • Mardi Gras (9 Feb 2016) 31,572
  • Valentine’s Day (13-15 Feb 2016) 76,013
  • The Oscars (27-28 Feb 2016) 55,680
  • Leap Year Day (29 Feb 2016) 28,806
  • St. Patrick’s Day (16-18 Mar 2016) 77,889
  • Easter (25-27 Mar 2016) 87,989
  • April Fool’s Day (31 Mar-1 Apr 2016) 102,341
  • Mother's Day (7-9 May 2016) 64,963

2. Tools + Assets

The process behind Collective Craving is adopted from Ben Fry's visualization framework that contains seven stages: acquire, parse, filter, mine, represent, refine, and interact. Tweets are acquired using TAGS by Martin Hawksey, then parsed and filtered using OpenRefine (formerly Google Refine). Next step is to do text mining and pre-processing using NLTK Library in Python and refine everything back on both OpenRefine & Excel. After a cleaner dataset is produced on the backend, it is being represented in HTML5/CSS/JS. The front-end of the interactive visualization is built using jQuery, D3.js and P5.js libraries. The main visualization is hugely inspired by Jim Vallandingham's awesome vintage bump chart.

For the digital creative assets, the food icon illustrations were all thanks to the ultra-talented couple Dita Wistarini (@ditut) and Pinot Ichwandardi (@pinot). Collective Craving's logo is made by Nitcha Fame Tothong, and the front-page parallax illustrations are made by Umi Syam.


Umi Syam is an interaction designer specializing in data art from Jakarta, Indonesia. She studied computer science at Bandung Institute of Technology and then worked for almost five years at Microsoft Indonesia as their front-end web gal and later jumped boat into a UX Evangelist. Collective Craving was developed as part of her thesis project for MFA Design & Technology, Parsons School of Design in 2016.

This project started as her own personal passion of investigating food culture amongst millennials. The reasoning? Clearly, because she's one of them.




See other data visualization projects.
Portfolio | LinkedIn | Instagram | Twitter

Collective Craving © 2016. Umi Syam. All Rights Reserved.