Name: Vivek Paul Joseph

Guide: Prof. Anirudha Joshi

Course: Interaction

Gamification of Corpus Cleaning

IDC maintains a steadily growing database of words for  regional languages. This database can be a resource to fuel development of new tools for the language. But this requires it to be error-free. Due to the complexity of the language, paired with its agglutinative property, it is a challenging to programmatically categorize the words.

The aim of this project is to crowdsource, through gamification, the tagging & correction of the words in the dataset. The end product was a mobile game that gamifies the dataset ‘cleaning’ activity. The product was designed, detailed and developed into a functioning prototype. Evaluation (with 14 users) indicated the system to be able to reliably clean the dataset.

How does it work?
– Users play the game (multiple game modes)
– They score points, stats, levels and try to climb to the top of the leaderboards.
– They compete other users in the multiplayer modes
– While they do all of the above, they help generate an error-free word database