Research map / Graph learning / Graph self-supervised learning
Multi-View Graph Representation for Programming Language Processing: An Investigation into Algorithm Detection
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use multi-view graph representations combining data flow, control flow, read-write interactions, and a combined graph | algorithm detection · POJ-104 | Accuracy (%): 94.83 -> 94.96 | Helped |
| Use multi-view graph representations instead of the strongest baseline LRPG | algorithm detection · ALG-109 | Micro-F1 (%): 60.56 $\pm$ 0.87 -> 65.26 $\pm$ 0.85 | Helped |
| Use multi-view graph representations instead of the strongest baseline LRPG | algorithm detection · ALG-10 | Micro-F1 (%): 78.48 $\pm$ 1.51 -> 80.15 $\pm$ 0.86 | Helped |
| Add the DFG view and its data-flow information to CG (ablation) | MVG · ALG-109 | Micro-F1 (%): 62.34 $\pm$ 1.11 -> 65.26 $\pm$ 0.85 | Helped |
| Add the CFG view and its control-flow information to CG (ablation) | MVG · ALG-109 | Micro-F1 (%): 64.18 $\pm$ 0.86 -> 65.26 $\pm$ 0.85 | Helped |
| Add the RWG view and its read-write information to CG (ablation) | MVG · ALG-109 | Micro-F1 (%): 64.01 $\pm$ 1.06 -> 65.26 $\pm$ 0.85 | Helped |
| Add the combined CG view alongside independent graph views (ablation) | MVG · ALG-109 | Micro-F1 (%): 64.38 $\pm$ 0.78 -> 65.26 $\pm$ 0.85 | Helped |
| Use separate DFG, CFG, RWG, and CG views instead of only CG (ablation) | MVG · ALG-109 | Micro-F1 (%): 62.02 $\pm$ 0.74 -> 65.26 $\pm$ 0.85 | Helped |
| Add an AST view (ablation) | MVG · ALG-109 | Micro-F1 (%): 65.26 $\pm$ 0.85 -> 65.19 $\pm$ 0.94 | Hurt |
Closest papers by meaning
- Global Relational Models of Source Code (ICLR 2020 · similarity 0.88)
- Learning to Represent Programs with Graphs (ICLR 2018 · similarity 0.87)
- Neural Code Comprehension: A Learnable Representation of Code Semantics (NeurIPS 2018 · similarity 0.87)
- GraphCodeBERT: Pre-training Code Representations with Data Flow (ICLR 2021 · similarity 0.87)
- Language-Agnostic Representation Learning of Source Code from Structure and Context (ICLR 2021 · similarity 0.86)
- JanusCoder: Towards a Foundational Visual-Programmatic Interface for Code Intelligence (ICLR 2026 · 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.