AI Paper Atlas

Research map / Graph learning / Knowledge and temporal graphs

Towards Better Evaluation for Dynamic Link Prediction

Farimah Poursafaei, Andy Huang, Kellin Pelrine, Reihaneh Rabbany

NeurIPS 2022 · Graph learning › Knowledge and temporal graphs · approach: Continuous-time dynamic graph learning · 30 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
switch evaluation from random to historical negative samplingCAWN · LastFMAU-ROC: 0.97 -> 0.40Hurt
switch evaluation from random to historical negative samplingCAWN · EnronAU-ROC: 0.93 -> 0.51Hurt
switch evaluation from random to historical negative samplingEdgeBank_inf · WikipediaAU-ROC: 0.91 -> 0.49Hurt
switch evaluation from random to historical negative samplingEdgeBank_tw · UN TradeAU-ROC: 0.67 -> 0.81Improved
switch evaluation from random to historical negative samplingTGN · FlightsAU-ROC: 0.80 -> 0.66Hurt
use time-window memory instead of storing all edges (EdgeBank_tw vs EdgeBank_inf)EdgeBank · historical NS, WikipediaAU-ROC: 0.49 -> 0.77Improved
use EdgeBank memorization instead of CAWN under random NSdynamic link prediction · LastFMAU-ROC: 0.97 (CAWN) vs 0.84 (EdgeBank_tw)Hurt
compare models by degree of correlation with EdgeBank_infmemorization-reliant models · historical and inductive NSperformance dropHurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy