Research map / Learning theory and algorithms / Prediction and decision losses
On the Relationship Between Relevance and Conflict in Online Social Link Recommendations
Yanbang Wang, Jon Kleinberg
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 |
|---|---|---|---|
| increase negative sampling rate eta from 1 toward 10 | most link recommendation algorithms · Reddit and Twitter | conflict awareness | Hurt |
| use GNN-based recommenders instead of heuristic ones as eta grows | link recommendation · Reddit and Twitter | recall | Improved |
| R-GCN versus Jaccard Index recommender | link recommendation · Reddit | conflict awareness: 0.2 -> 0.95 | Improved |
| use conflict minimization solver that ignores relevance | link recommendation · Reddit and Twitter | recall | Hurt |
| use conflict minimization solver that ignores relevance | link recommendation · Reddit and Twitter | conflict awareness | Improved |
| add a single link to the network | Friedkin-Johnsen opinion conflict · ER, path, grid, Karate, Reddit, Twitter | conflict change | Improved |
| add edges one by one up to a complete graph | conflict contraction rate · N=20 nodes | contraction rate vs bounds | No clear effect |
Closest papers by meaning
- Optimizing Social Network Interventions via Hypergradient-Based Recommender System Design (ICML 2025 · similarity 0.85)
- Reaching a Consensus in Predictive Loops (NeurIPS 2026 · similarity 0.84)
- The Burden of Interactive Alignment with Inconsistent Preferences (NeurIPS 2025 · similarity 0.83)
- Networked Inequality: Preferential Attachment Bias in Graph Neural Network Link Prediction (ICML 2024 · similarity 0.82)
- Representative Ranking for Deliberation in the Public Sphere (ICML 2025 · similarity 0.82)
- Revisiting Link Prediction: a data perspective (ICLR 2024 · similarity 0.82)
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.
