Research map / Learning theory and algorithms / Optimal transport
IdeFN: Identifying Unclicked Space False Negatives via Relaxed Partial Optimal Transport for Conversion Rate Prediction
Weiyi Zhong, Weiming Liu, Lianyong Qi, Xiaoran Zhao, Xiaolong Xu, Haolong Xiang, Yang Cao, Shichao Pei, Qiang Ni
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add IdeFN false-negative relabeling to ESMM | ESMM (MLP) · Ali-CCP | AUC: 0.6313 -> 0.6656 | Improved |
| add IdeFN to MMoE | MMoE · Ali-CCP | AUC: 0.6241 -> 0.6641 | Improved |
| add IdeFN to MMoE | MMoE · KuaiRand-Pure | AUC: 0.8310 -> 0.8672 | Improved |
| add IdeFN to ESMM | ESMM (MLP) · KuaiRand-Pure | LogLoss: 0.0703 -> 0.0632 | Improved |
| IdeFN vs runner-up NISE | IdeFN (MLP) · Ali-CCP | AUC: 0.6515 -> 0.6656 | Improved |
| IdeFN vs runner-up NISE | IdeFN (DCNV2) · KuaiRand-Pure | AUC: 0.8588 -> 0.8756 | Improved |
| ESCM2-DR with DCNV2 backbone | ESCM2-DR · KuaiRand-Pure | AUC: 0.8101 (DeepFM) -> 0.6442 (DCNV2) | Hurt |
| add CTR auxiliary task (ablation) | IdeFN · Ali-CCP | AUC/KS | Improved |
| add RPOT module (identify false negatives in unclicked space) (ablation) | IdeFN · Ali-CCP | AUC/KS | Improved |
| increase kappa (fraction of unclicked allowed to match positives) too large (ablation) | IdeFN · KuaiRand-Pure | AUC/KS | Hurt |
| vary CVR/CTR loss weight gamma away from 30 (ablation) | IdeFN · KuaiRand-Pure | AUC/KS | Hurt |
| vary alpha away from 0.8 (ablation) | IdeFN · KuaiRand-Pure | AUC/KS | Hurt |
Closest papers by meaning
- Entire-Space Variational Information Exploitation for Post-Click Conversion Rate Prediction (AAAI 2025 · similarity 0.92)
- Online Conversion Rate Prediction via Multi-Interval Screening and Synthesizing under Delayed Feedback (AAAI 2024 · similarity 0.86)
- AT4CTR: Auxiliary Match Tasks for Enhancing Click-Through Rate Prediction (AAAI 2024 · similarity 0.86)
- Addressing Cold-Start Problem in Click-Through Rate Prediction via Supervised Diffusion Modeling (AAAI 2025 · similarity 0.84)
- From Feature Interaction to Feature Generation: A Generative Paradigm of CTR Prediction Models (ICML 2025 · similarity 0.83)
- Adaptive Hardness Negative Sampling for Collaborative Filtering (AAAI 2024 · similarity 0.83)
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.
