Group recommender systems produce suggestions in contexts in which more than one person is involved in the recommendation process. They present additional tasks w.r.t. those for single users, such as the identification of the groups, or their modeling. While this clearly amplifies the possible reproducibility issues, to date, no framework to benchmark group recommender systems exists.

In a study published in the proceedings of the CCIA 2022 conference, with Joaquin Dario Silveira and Maria Salamó, we enable reproducibility in group recommender systems by extending the LibRec library, which stands out as one of the richest, with more than 70 different recommender algorithms, good performance and several evaluation metrics. Specifically, we include several approaches for all the stages of group recommender systems: group formation, group modeling strategies, and evaluation.

The framework is available at: https://github.com/panserbjorn/librec/tree/3.0.0/RecSys.

Group recommendation pipeline

The previous figure depicts the recommendation pipeline for GroupLibRec, which includes the LibRec pipeline (elements in white color). Note that the LibRec library pipeline performs the steps of data ingestion, splitting it into train and test sets, training a recommender, generating recommendations, and, finally, evaluating the results. Configuration files enable the setup of all the required parameters in a run, such as the recommendation algorithm selection, the dataset location, and the evaluation metrics. Thus, this process facilitates the reproducibility of experiments just by sharing and exchanging a single configuration file.

The red components in the figure show the extension of the pipeline for enabling group recommendations. The additional steps are: group building, group modeling, and splitting. Specifically, the Group Recommender delegates the responsibility of generating the individual recommendations in the pipeline to the individual user recommender algorithm and then it builds the recommendations for the groups. Specific additions to the recommendation process for enabling group recommendations will be discussed in the following sections.