AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

CoLink: An Unsupervised Framework for User Identity Linkage

Zexuan Zhong, Yong Cao, Yong Cao, Mu Guo, Mu Guo, Zaiqing Nie, Zaiqing Nie

AAAI 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 68 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
co-training with seq2seq attribute model versus Alias-disambCoLink · Enterprise-to-LinkedIn UIL (594 linked pairs)F1: 68.69 -> 89.02Improved
replace SVM attribute model with sequence-to-sequence (ablation)CoLink (Fine-tuned) · Enterprise-to-LinkedIn UIL (594 linked pairs)F1: 71.81 -> 89.02Improved
add final revision after co-training convergence (coarse seeds) (ablation)CoLink (S2S+Coarse-tuned) · Enterprise-to-LinkedIn UIL (594 linked pairs)F1: 85.18 -> 87.82Improved
add final revision after co-training convergence (fine seeds) (ablation)CoLink (S2S+Fine-tuned) · Enterprise-to-LinkedIn UIL (594 linked pairs)F1: 89.02 -> 87.77Hurt
use coarse-tuned (81 pairs, biased) instead of fine-tuned seed rules (ablation)CoLink (S2S) · Enterprise-to-LinkedIn UIL (594 linked pairs)F1: 89.02 -> 85.18Hurt
20% noisy seed pairs (ablation)CoLink (S2S) · Enterprise-to-LinkedIn UIL (594 linked pairs)F1No clear effect
stricter likelihood threshold (smaller percentage) (ablation)CoLink (S2S) · Enterprise-to-LinkedIn UIL (594 linked pairs)precision/recallMixed

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