Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -27,19 +27,22 @@ We evaluated our model against traditional heuristics, machine learning approach
|
|
| 27 |
<img src="analysis/overall_performance_comparison.png" alt="Model Comparison" width="100%"/>
|
| 28 |
</div>
|
| 29 |
|
| 30 |
-
| Model | Kendall's τ |
|
| 31 |
-
|
| 32 |
-
| **Chunkr Reading Order** | **
|
| 33 |
-
| openai/gpt-5 |
|
| 34 |
-
|
|
| 35 |
-
|
|
| 36 |
-
| Pairwise
|
| 37 |
-
|
|
| 38 |
-
| anthropic/claude-sonnet-4.5 |
|
| 39 |
-
|
|
| 40 |
-
|
|
| 41 |
-
|
|
| 42 |
-
| z-ai/glm-4.5v |
|
|
|
|
|
|
|
|
|
|
| 43 |
|
| 44 |
## Dataset Overview
|
| 45 |
|
|
|
|
| 27 |
<img src="analysis/overall_performance_comparison.png" alt="Model Comparison" width="100%"/>
|
| 28 |
</div>
|
| 29 |
|
| 30 |
+
| Model | Exact Match | Kendall's τ | Spearman's ρ | Position Acc | Inference (ms) |
|
| 31 |
+
|-------|------------|-------------|--------------|--------------|----------------|
|
| 32 |
+
| **Chunkr Reading Order** | **88.1%** | **98.2%** | **98.5%** | **96.1%** | **44.6** |
|
| 33 |
+
| openai/gpt-5 | 74.8% | 85.0% | 85.3% | 82.5% | 15035.2 |
|
| 34 |
+
| google/gemini-2.5-pro | 72.4% | 82.1% | 82.5% | 79.3% | 4678.5 |
|
| 35 |
+
| x-ai/grok-4-fast | 68.8% | 91.4% | 92.5% | 83.2% | 5661.1 |
|
| 36 |
+
| Pairwise XGBoost* | 67.7% | 92.8% | 94.6% | 84.7% | 16.7 |
|
| 37 |
+
| Pairwise LightGBM* | 67.4% | 92.9% | 94.9% | 84.3% | 33.6 |
|
| 38 |
+
| anthropic/claude-sonnet-4.5 | 53.8% | 84.5% | 87.1% | 73.1% | 919.4 |
|
| 39 |
+
| XY-Cut | 51.7% | 84.4% | 85.5% | 68.5% | 1.7 |
|
| 40 |
+
| TBLR (Baseline) | 50.3% | 87.5% | 90.4% | 75.2% | 1.7 |
|
| 41 |
+
| qwen/qwen3-vl-235b-a22b-instruct | 49.1% | 79.7% | 80.9% | 66.7% | 4668.5 |
|
| 42 |
+
| z-ai/glm-4.5v | 46.8% | 74.5% | 75.4% | 61.0% | 8968.0 |
|
| 43 |
+
| LayoutLMv3 Base | 9.4% | 12.0% | 13.0% | 23.9% | 44.2 |
|
| 44 |
+
|
| 45 |
+
\*Pairwise models trained on Chunkr dataset with optimized features
|
| 46 |
|
| 47 |
## Dataset Overview
|
| 48 |
|