SwapIt: a Recommender System for Exchange Platforms

Jérémie Rappaz, Maria Vladarean
Julian McAuley, Michele Catasta


Welcome to the SwapIt project! We are a team of students and researchers from Ecole Polytechnique Fédérale de Lausanne (EPFL) and UC San Diego who are passionate about recommender systems and alternative economic models. In our quest to build a recommender system for bartering communities, we analysed the data of several bartering platforms covering beers, books and video games. We designed an approach to recommending trades which takes into consideration the peculiarities of the data, and yields promising results on 3 distinct bartering datasets.

If you want to know more:   Read the WSDM2017 paper describing our approach!
  Take a look at the poster


  Check out the SwapIt recommender code (Matlab)   Download a .zip file of the datasets   View the slides from our talk at WSDM2017