Implicit degree bias in the link prediction task
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866929363376668672 |
|---|---|
| author | Aiyappa, Rachith Wang, Xin Kim, Munjung Seckin, Ozgur Can Yoon, Jisung Ahn, Yong-Yeol Kojaku, Sadamori |
| author_facet | Aiyappa, Rachith Wang, Xin Kim, Munjung Seckin, Ozgur Can Yoon, Jisung Ahn, Yong-Yeol Kojaku, Sadamori |
| contents | Link prediction -- a task of distinguishing actual hidden edges from random unconnected node pairs -- is one of the quintessential tasks in graph machine learning. Despite being widely accepted as a universal benchmark and a downstream task for representation learning, the validity of the link prediction benchmark itself has been rarely questioned. Here, we show that the common edge sampling procedure in the link prediction task has an implicit bias toward high-degree nodes and produces a highly skewed evaluation that favors methods overly dependent on node degree, to the extent that a ``null'' link prediction method based solely on node degree can yield nearly optimal performance. We propose a degree-corrected link prediction task that offers a more reasonable assessment that aligns better with the performance in the recommendation task. Finally, we demonstrate that the degree-corrected benchmark can more effectively train graph machine-learning models by reducing overfitting to node degrees and facilitating the learning of relevant structures in graphs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_14985 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Implicit degree bias in the link prediction task Aiyappa, Rachith Wang, Xin Kim, Munjung Seckin, Ozgur Can Yoon, Jisung Ahn, Yong-Yeol Kojaku, Sadamori Social and Information Networks Physics and Society Link prediction -- a task of distinguishing actual hidden edges from random unconnected node pairs -- is one of the quintessential tasks in graph machine learning. Despite being widely accepted as a universal benchmark and a downstream task for representation learning, the validity of the link prediction benchmark itself has been rarely questioned. Here, we show that the common edge sampling procedure in the link prediction task has an implicit bias toward high-degree nodes and produces a highly skewed evaluation that favors methods overly dependent on node degree, to the extent that a ``null'' link prediction method based solely on node degree can yield nearly optimal performance. We propose a degree-corrected link prediction task that offers a more reasonable assessment that aligns better with the performance in the recommendation task. Finally, we demonstrate that the degree-corrected benchmark can more effectively train graph machine-learning models by reducing overfitting to node degrees and facilitating the learning of relevant structures in graphs. |
| title | Implicit degree bias in the link prediction task |
| topic | Social and Information Networks Physics and Society |
| url | https://arxiv.org/abs/2405.14985 |