Papers › Strengthening structural baselines for graph classification using Local Topological Profile
Strengthening structural baselines for graph classification using Local Topological Profile
Jakub Adamczyk, Wojciech Czech
We present the analysis of the topological graph descriptor Local Degree Profile (LDP), which forms a widely used structural baseline for graph classification. Our study focuses on model evaluation in the context of the recently developed fair evaluation framework, which defines rigorous routines for model selection and evaluation for graph classification, ensuring reproducibility and comparability of the results. Based on the obtained insights, we propose a new baseline algorithm called Local Topological Profile (LTP), which extends LDP by using additional centrality measures and local vertex descriptors. The new approach provides the results outperforming or very close to the latest GNNs for all datasets used. Specifically, state-of-the-art results were obtained for 4 out of 9 benchmark datasets. We also consider computational aspects of LDP-based feature extraction and model construction to propose practical improvements affecting execution speed and scalability. This allows for handling modern, large datasets and extends the portfolio of benchmarks used in graph representation learning. As the outcome of our work, we obtained LTP as a simple to understand, fast and scalable, still robust baseline, capable of outcompeting modern graph classification models such as Graph Isomorphism Network (GIN). We provide open-source implementation at \href{https://github.com/j-adamczyk/LTP}{GitHub}.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Graph Classification | COLLAB | Local Topological Profile (LTP) | Accuracy | 79.4 ± 2.5 | #17 of 39 | Archive leaderboard | report |
| Graph Classification | COLLAB | Local Topological Profile (LTP) | Accuracy (10-fold) | 79.4 ± 2.5 | #17 of 39 | Archive leaderboard | report |
| Graph Classification | ENZYMES | Local Topological Profile (LTP) | Accuracy | 42.5 ± 4.1 | #48 of 54 | Archive leaderboard | report |
| Graph Classification | ENZYMES | Local Topological Profile (LTP) | Accuracy (10-fold) | 42.5 ± 4.1 | #48 of 54 | Archive leaderboard | report |
| Graph Classification | IMDB-BINARY | Local Topological Profile (LTP) | Accuracy | 74.5 ± 4.3 | #1 of 8 | Archive leaderboard | report |
| Graph Classification | IMDB-BINARY | Local Topological Profile (LTP) | Accuracy (10-fold) | 74.5 ± 4.3 | #1 of 8 | Archive leaderboard | report |
| Graph Classification | IMDB-MULTI | Local Topological Profile (LTP) | Accuracy | 50.0 ± 4.6 | #1 of 1 | Archive leaderboard | report |
| Graph Classification | IMDB-MULTI | Local Topological Profile (LTP) | Accuracy (10-fold) | 50.0 ± 4.6 | #1 of 1 | Archive leaderboard | report |
| Graph Classification | IMDb-B | Local Topological Profile (LTP) | Accuracy | 74.5 ± 4.3 | #26 of 51 | Archive leaderboard | report |
| Graph Classification | IMDb-B | Local Topological Profile (LTP) | Accuracy (10-fold) | 74.5 ± 4.3 | #26 of 51 | Archive leaderboard | report |
| Graph Classification | IMDb-M | Local Topological Profile (LTP) | Accuracy | 50.0 ± 4.6 | #24 of 36 | Archive leaderboard | report |
| Graph Classification | IMDb-M | Local Topological Profile (LTP) | Accuracy (10-fold) | 50.0 ± 4.6 | #24 of 36 | Archive leaderboard | report |
| Graph Classification | NCI1 | Local Topological Profile (LTP) | Accuracy | 77.1 ± 3.7 | #43 of 69 | Archive leaderboard | report |
| Graph Classification | NCI1 | Local Topological Profile (LTP) | Accuracy (10-fold) | 77.1 ± 3.7 | #43 of 69 | Archive leaderboard | report |
| Graph Classification | PROTEINS | Local Topological Profile (LTP) | Accuracy | 72.7 ± 4.2 | #92 of 103 | Archive leaderboard | report |
| Graph Classification | PROTEINS | Local Topological Profile (LTP) | Accuracy (10 fold) | 72.7 ± 4.2 | #92 of 103 | Archive leaderboard | report |
| Graph Classification | REDDIT-B | Local Topological Profile (LTP) | Accuracy | 91.1 ± 1.0 | #6 of 12 | Archive leaderboard | report |
| Graph Classification | REDDIT-B | Local Topological Profile (LTP) | Accuracy (10-fold) | 91.1 ± 1.0 | #6 of 12 | Archive leaderboard | report |
| Graph Classification | REDDIT-BINARY | Local Topological Profile (LTP) | Accuracy | 91.1 ± 1.0 | #2 of 9 | Archive leaderboard | report |
| Graph Classification | REDDIT-BINARY | Local Topological Profile (LTP) | Accuracy (10-fold) | 91.1 ± 1.0 | #2 of 9 | Archive leaderboard | report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Methods
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections