Research map / Graph learning / Knowledge and temporal graphs
Efficiently Answering Technical Questions — A Knowledge Graph Approach
Shuo Yang, Lei Zou, Zhongyuan Wang, Jun Yan, Ji-Rong Wen
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 |
|---|---|---|---|
| re-rank STEP-1 candidates with knowledge-graph random walk (ablation) | question-driven retrieval · EVAL1 | A@5: 0.31 -> 0.46 | Improved |
| re-rank STEP-1 candidates with knowledge-graph random walk (ablation) | question-driven retrieval · EVAL2 | A@5: 0.44 -> 0.61 | Improved |
| question-driven candidate selection (via similar logged questions) instead of document-driven retrieval | candidate selection · EVAL1 | A@100: 0.32 -> 0.87 | Improved |
| KG re-ranking vs best baseline LETOR(1:3) | technical QA re-ranking · EVAL1 | A@5: 0.42 -> 0.46 | Improved |
| KG re-ranking vs best baseline LETOR(1:3) | technical QA re-ranking · EVAL1 | A@1: 0.21 -> 0.22 | No clear effect |
| increase random walk iterations (sampling) up to 4M | random-walk re-ranking · EVAL1 | A@1/A@3/A@5 | Improved |
| index-based random walk with materialized nodes instead of sampling | online search · response time | time: ~7 minutes -> 100-800 milliseconds | Improved |
| raise controlled path length l (ablation) | index of materialized nodes · index size | materialized nodes: ~40,000 nodes at l=20 | Improved |
Closest papers by meaning
- Harnessing Large Language Models for Knowledge Graph Question Answering via Adaptive Multi-Aspect Retrieval-Augmentation (AAAI 2025 · similarity 0.86)
- Knowledge Graph Prompting for Multi-Document Question Answering (AAAI 2024 · similarity 0.86)
- Simple is Effective: The Roles of Graphs and Large Language Models in Knowledge-Graph-Based Retrieval-Augmented Generation (ICLR 2025 · similarity 0.85)
- Think-on-Graph 2.0: Deep and Faithful Large Language Model Reasoning with Knowledge-guided Retrieval Augmented Generation (ICLR 2025 · similarity 0.85)
- BRIGHT: A Realistic and Challenging Benchmark for Reasoning-Intensive Retrieval (ICLR 2025 · similarity 0.84)
- Knowledge Editing with Dynamic Knowledge Graphs for Multi-Hop Question Answering (AAAI 2025 · similarity 0.84)
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.
