Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -492,26 +492,6 @@ RevUtil focuses on four main aspects of review comments:
|
|
| 492 |
|
| 493 |
---
|
| 494 |
|
| 495 |
-
## 🤖 RevUtil Synthetic
|
| 496 |
-
|
| 497 |
-
- **10,000** synthetically labeled review comments generated with **GPT-4o**.
|
| 498 |
-
- Each example includes both a **score** and a **rationale**.
|
| 499 |
-
- Split into **9k train** / **1k test**.
|
| 500 |
-
|
| 501 |
-
**Key columns:**
|
| 502 |
-
|
| 503 |
-
| Column | Description |
|
| 504 |
-
| -------------------------- | -------------------------------------------- |
|
| 505 |
-
| `paper_id` | ID of the reviewed paper |
|
| 506 |
-
| `venue` | Conference or journal name |
|
| 507 |
-
| `focused_review` | Full review (weakness + suggestion sections) |
|
| 508 |
-
| `review_point` | Individual review comment |
|
| 509 |
-
| `id` | Unique ID for the review point |
|
| 510 |
-
| `chatgpt_ASPECT_score` | Model-generated score for the aspect |
|
| 511 |
-
| `chatgpt_ASPECT_rationale` | Explanation of the score provided by GPT-4o |
|
| 512 |
-
|
| 513 |
-
---
|
| 514 |
-
|
| 515 |
## 🚀 Usage
|
| 516 |
|
| 517 |
You can load the datasets directly via 🤗 Datasets:
|
|
|
|
| 492 |
|
| 493 |
---
|
| 494 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 495 |
## 🚀 Usage
|
| 496 |
|
| 497 |
You can load the datasets directly via 🤗 Datasets:
|