Beyond-accuracy: a review on diversity, serendipity, and fairness in recommender systems based on graph neural networks

Abstract

By providing personalized suggestions to users, recommender systems have become essential to numerous online platforms. Collaborative filtering, particularly graph-based approaches using Graph Neural Networks (GNNs), have demonstrated great results in terms of recommendation accuracy. However, accuracy may not always be the most important criterion for evaluating recommender systems' performance, since beyond-accuracy aspects such as recommendation diversity, serendipity, and fairness can strongly influence user engagement and satisfaction. This review paper focuses on addressing these dimensions in GNN-based recommender systems, going beyond the conventional accuracy-centric perspective. We begin by reviewing recent developments in approaches that improve not only the accuracy-diversity trade-off but also promote serendipity, and fairness in GNN-based recommender systems. We discuss different stages of model development including data preprocessing, graph construction, embedding initialization, propagation layers, embedding fusion, score computation, and training methodologies. Furthermore, we present a look into the practical difficulties encountered in assuring diversity, serendipity, and fairness, while retaining high accuracy. Finally, we discuss potential future research directions for developing more robust GNN-based recommender systems that go beyond the unidimensional perspective of focusing solely on accuracy. This review aims to provide researchers and practitioners with an in-depth understanding of the multifaceted issues that arise when designing GNN-based recommender systems, setting our work apart by offering a comprehensive exploration of beyond-accuracy dimensions.

Future-Present Learning in Place: postdigital learning at the scale of the city

“…Recommender systems as knowledge delivery mediators therefore become integral to interacting with the knowledge web. Subsequent following sections focus on recommender systems in the potentially required conditions of anonymised user data, privacy protection, and the increased understanding of the importance of serendipity as a part of search result suggestions [58]. The roles that Large Language Model 26 trained artificial intelligence (AI) tools might play are considered, particularly in light of possible impact on the integrity of the knowledge web.…”

“…Search result recommender systems that might be best employed in a CLN would most likely be based on anonymised user profile interactions and employ a system whereby related topic results could be offered that include opportunity for surprise and further exploration. Duricic et al [58] emphasise that "accuracy may not always be the most important criterion" of graph neural network (GNN) 27 based systems, as "aspects such as recommendation diversity, serendipity, and fairness can strongly influence user engagement and satisfaction". Within the context of diversity (of content) and fairness (of returned results against others), serendipity would seem to be of most significance for CLN recommendations, as it "indicates the unexpected nature of recommendations, (and) encourages users to explore beyond their usual preferences and stimulates curiosity" [58].…”

“…Duricic et al [58] emphasise that "accuracy may not always be the most important criterion" of graph neural network (GNN) 27 based systems, as "aspects such as recommendation diversity, serendipity, and fairness can strongly influence user engagement and satisfaction". Within the context of diversity (of content) and fairness (of returned results against others), serendipity would seem to be of most significance for CLN recommendations, as it "indicates the unexpected nature of recommendations, (and) encourages users to explore beyond their usual preferences and stimulates curiosity" [58]. GNNs use "collaborative filtering" techniques, that begin with data preprocessing of "user-item interaction data and auxiliary information such as user/item features or social connections" [58].…”

“…Within the context of diversity (of content) and fairness (of returned results against others), serendipity would seem to be of most significance for CLN recommendations, as it "indicates the unexpected nature of recommendations, (and) encourages users to explore beyond their usual preferences and stimulates curiosity" [58]. GNNs use "collaborative filtering" techniques, that begin with data preprocessing of "user-item interaction data and auxiliary information such as user/item features or social connections" [58]. 28 .…”