Post Processing Recommender Systems with Knowledge Graphs for Recency, Popularity, and Diversity of Explanations

24 Apr 2022  ยท  Giacomo Balloccu, Ludovico Boratto, Gianni Fenu, Mirko Marras ยท

Existing explainable recommender systems have mainly modeled relationships between recommended and already experienced products, and shaped explanation types accordingly (e.g., movie "x" starred by actress "y" recommended to a user because that user watched other movies with "y" as an actress). However, none of these systems has investigated the extent to which properties of a single explanation (e.g., the recency of interaction with that actress) and of a group of explanations for a recommended list (e.g., the diversity of the explanation types) can influence the perceived explaination quality. In this paper, we conceptualized three novel properties that model the quality of the explanations (linking interaction recency, shared entity popularity, and explanation type diversity) and proposed re-ranking approaches able to optimize for these properties. Experiments on two public data sets showed that our approaches can increase explanation quality according to the proposed properties, fairly across demographic groups, while preserving recommendation utility. The source code and data are available at https://github.com/giacoballoccu/explanation-quality-recsys.

PDF Abstract

Datasets


Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Music Recommendation Last.FM FM NDCG 0.15 # 1
Explainable Recommendation Last.FM DPR-PGPR NDCG 0.13 # 2
Music Recommendation Last.FM PGPR NDCG 0.14 # 2
Music Recommendation Last.FM NMF NDCG 0.14 # 2
Explainable Recommendation Last.FM P-PGPR NDCG 0.13 # 2
Music Recommendation Last.FM BPR NDCG 0.13 # 4
Explainable Recommendation Last.FM R-PGPR NDCG 0.14 # 1
Explainable Recommendation MovieLens 1M D-PGPR NDCG 0.32 # 3
Movie Recommendation MovieLens 1M BPR NDCG 0.33 # 1
Movie Recommendation MovieLens 1M FM NDCG 0.32 # 3
Explainable Recommendation MovieLens 1M P-PGPR NDCG 0.33 # 2
Explainable Recommendation MovieLens 1M DPR-PGPR NDCG 0.31 # 4
Explainable Recommendation MovieLens 1M R-PGPR NDCG 0.34 # 1
Movie Recommendation MovieLens 1M KGAT NDCG 0.33 # 1
Movie Recommendation MovieLens 1M CFKG NDCG 0.27 # 4
Explainable Recommendation MovieLens 1M DP-PGPR NDCG 0.31 # 4

Methods