Datasets to Evaluate Accuracy, Miscalibration and Popularity Lift in Recommendations

<p>This repository contains three datasets for evaluating accuracy, miscalibration and popularity lift in recommender systems. All datasets contain genre/category information in addition to different user group splits:</p> <ol> <li>Last.fm (lfm.zip), based on the LFM-1b dataset of JKU Linz (http://www.cp.jku.at/datasets/LFM-1b/)</li> <li>MovieLens (ml.zip), based on MovieLens-1M dataset (https://grouplens.org/datasets/movielens/1m/)</li> <li>MyAnimeList (anime.zip), based on the MyAnimeList dataset of Kaggle (https://www.kaggle.com/CooperUnion/anime-recommendations-database)</li> </ol> <p>&#39;user_events_cats.txt&#39; contains the users&#39; rating/interaction data along with a list of genres/categories assigend to the rated items. The list of categories is given in &#39;categories.txt&#39;. Additionally, assignments to three user groups that differ in their inclination to popular/mainstream items are provided: LowPop in &#39;low_main_users.txt&#39;, MedPop in &#39;med_main_users.txt&#39;, and HighPop in &#39;high_main_users.txt&#39;.</p> <p>The format of the three user files are &quot;user,mainstreaminess&quot;</p> <p>The format of the user-events files are &quot;user,item,preference,cats&quot;, where different categories are separated by &#39;|&#39;</p> <p>The format of the categories files are &quot;category-name,index&quot;, where index refers to the category-id in the user-events files</p> <p>Example Python-code for analyzing the datasets as well as empirical results on calibration, popularity lift and accuracy can be found on GitHub: https://github.com/domkowald/FairRecSys</p>