WTD, a sampling approach to debiasing the offline evaluation of Recommender Systems

This code is related to the following publications:

1. Chapter 3 of the PhD thesis,: "Active Learning in RecommenderSystems: An Unbiased and Beyond-Accuracy Perspective", by Diego Carraro. (to be published)
2. Journal article: "A Sampling Approach to Debiasing the Offline Evaluation of Recommender Systems", by Diego Carraro and Derek Bridge. (to be published)
3. Conference article: "Debiased Offline Evaluation of Recommender Systems: A Weighted-Sampling Approach", by Diego Carraro and Derek Bridge. Procs. of the 35th Annual ACM Symposium on Applied Computing, ACM, pp.1435-1442, 2020.

The repository is divided into two Jupiter notebooks:

1. "Debiasing Intervention.ipynb", which provides the code to reproduce the data preparation for the experiments performed in the various publications.
2. "Create Sample Dataset.ipynb", which creates sample data to test the code. Indeed, datasests used for the experiments are not pubicly available, but only available under request