Sentimentalamp is a floating IoT lamp suspended by magnets. The temperature of the color is determined by the average Twitter sentiment within 50 miles, as determined by an NLP sentiment analysis tool. This was an individual project that I completed as part of the MITidm core curriculum.

  • Product Design
  • MITidm
  • Natural Language Processing
  • Internet of Things
  • Design Challenge
  • The prompt given was to design a lamp of any kind. Given my interest in the Internet of Things and AI/ML, I wanted to design something that leveraged information from the internet and gave a useful or interesting display of that information in the form of light.
  • Project Scope
  • I took this project from inception through many rounds of iteration (making something float is very hard, it turns out) to the version seen on this page. At the time of this writing, I consider the project to be partially incomplete. I was able to finish all of the functionality but not spend nearly any time on industrial design. I did not write the Naive Bayes Twitter sentiment analysis tool from scratch, but followed a walkthrough which can be found here.
Making abstract information like Twitter sentiment visible can make it more tangible and allow us to be more in touch with our community.

The internet, despite the incredible breadth of information available, has few data streams that are immediately and viscerally useful to us. This was my attempt to experiment with ways we might change that.

See the Github Repository
  • Target UserCommunity Oriented Digital Natives
  • Type of WorkProduct Design, Coding, Electronics
  • Tools UsedPython, C++, Arduino, Naive Bayes Classifier, 3D Printing, Neopixel
  • LocationCambridge, MA