RanKit is an interactive prototype ranking tool for building your own rankings based on a small set of input preferences. Rnkit offers 3 types of interaction: build a sublist of items, group items into categories, or compare pairs of items to give the system a sample of your preferences. Then a ranking over all the items in the dataset is automatically generated. You can try out the tool using a number of sample data sets including colleges, movies, and US states.

The Rankit System was developed and evaluated by a team of undergrad students at WPI for their Major Qualifying Project under the supervision of Caitlin Kuhlman.

Our 2018 CIKM demo paper describes the system in detail and provides case studies for personalized ranking.

Our 2019 SIGCHI paper provides and in-depth evaluation of different ranking interaction modes using a crowdsourced user study and a theoretical analysis of the tradeoffs between user effort required and rank quality.