Enregistré dans:
Détails bibliographiques
Auteurs principaux: Cornell, Filip, Smirnov, Oleg, Gandler, Gabriela Zarzar, Cao, Lele
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2506.12588
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912429979467776
author Cornell, Filip
Smirnov, Oleg
Gandler, Gabriela Zarzar
Cao, Lele
author_facet Cornell, Filip
Smirnov, Oleg
Gandler, Gabriela Zarzar
Cao, Lele
contents Recent work has questioned the reliability of graph learning benchmarks, citing concerns around task design, methodological rigor, and data suitability. In this extended abstract, we contribute to this discussion by focusing on evaluation strategies in Temporal Link Prediction (TLP). We observe that current evaluation protocols are often affected by one or more of the following issues: (1) inconsistent sampled metrics, (2) reliance on hard negative sampling often introduced as a means to improve robustness, and (3) metrics that implicitly assume equal base probabilities across source nodes by combining predictions. We support these claims through illustrative examples and connections to longstanding concerns in the recommender systems community. Our ongoing work aims to systematically characterize these problems and explore alternatives that can lead to more robust and interpretable evaluation. We conclude with a discussion of potential directions for improving the reliability of TLP benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2506_12588
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Are We Really Measuring Progress? Transferring Insights from Evaluating Recommender Systems to Temporal Link Prediction
Cornell, Filip
Smirnov, Oleg
Gandler, Gabriela Zarzar
Cao, Lele
Machine Learning
Recent work has questioned the reliability of graph learning benchmarks, citing concerns around task design, methodological rigor, and data suitability. In this extended abstract, we contribute to this discussion by focusing on evaluation strategies in Temporal Link Prediction (TLP). We observe that current evaluation protocols are often affected by one or more of the following issues: (1) inconsistent sampled metrics, (2) reliance on hard negative sampling often introduced as a means to improve robustness, and (3) metrics that implicitly assume equal base probabilities across source nodes by combining predictions. We support these claims through illustrative examples and connections to longstanding concerns in the recommender systems community. Our ongoing work aims to systematically characterize these problems and explore alternatives that can lead to more robust and interpretable evaluation. We conclude with a discussion of potential directions for improving the reliability of TLP benchmarks.
title Are We Really Measuring Progress? Transferring Insights from Evaluating Recommender Systems to Temporal Link Prediction
topic Machine Learning
url https://arxiv.org/abs/2506.12588