Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -115,7 +115,6 @@ configs:
|
|
| 115 |
data_files:
|
| 116 |
- split: train
|
| 117 |
path: industry_prediction/train-*
|
| 118 |
-
license: cc-by-4.0
|
| 119 |
task_categories:
|
| 120 |
- text-classification
|
| 121 |
language:
|
|
@@ -133,6 +132,9 @@ size_categories:
|
|
| 133 |
EDINET-Bench is a Japanese financial benchmark designed to evaluate the performance of LLMs on challenging financial tasks including accounting fraud detection, earnings forecasting, and industry prediction.
|
| 134 |
This dataset is built leveraging [EDINET](https://disclosure2.edinet-fsa.go.jp), a platform managed by the Financial Services Agency (FSA) of Japan that provides access to disclosure documents such as securities reports.
|
| 135 |
|
|
|
|
|
|
|
|
|
|
| 136 |
## Resources
|
| 137 |
- 📃**Paper**: Read our paper for detailed dataset construction pipeline and evaluation results at https://pub.sakana.ai/edinet-bench
|
| 138 |
- 🏗️**Counstruction Code**: Create a new benchmark dataset at https://github.com/SakanaAI/edinet2dataset
|
|
@@ -176,7 +178,6 @@ DatasetDict({
|
|
| 176 |
})
|
| 177 |
```
|
| 178 |
|
| 179 |
-
Please see amended reports labeled as fraud cases at [Google Drive](https://drive.google.com/drive/folders/1ZD-AySiZJ66S2qS6kfZDTm0SnBIrCQp8?usp=sharing).
|
| 180 |
|
| 181 |
**Earnings forecast**
|
| 182 |
|
|
@@ -226,7 +227,7 @@ Future research directions could explore the development of benchmark task desig
|
|
| 226 |
|
| 227 |
## LICENSE
|
| 228 |
|
| 229 |
-
EDINET-Bench is licensed under the [
|
| 230 |
|
| 231 |
|
| 232 |
## ⚠️ Warnings
|
|
|
|
| 115 |
data_files:
|
| 116 |
- split: train
|
| 117 |
path: industry_prediction/train-*
|
|
|
|
| 118 |
task_categories:
|
| 119 |
- text-classification
|
| 120 |
language:
|
|
|
|
| 132 |
EDINET-Bench is a Japanese financial benchmark designed to evaluate the performance of LLMs on challenging financial tasks including accounting fraud detection, earnings forecasting, and industry prediction.
|
| 133 |
This dataset is built leveraging [EDINET](https://disclosure2.edinet-fsa.go.jp), a platform managed by the Financial Services Agency (FSA) of Japan that provides access to disclosure documents such as securities reports.
|
| 134 |
|
| 135 |
+
## Notice
|
| 136 |
+
- **June 9, 2025**: This dataset was originally released under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Although section 1.7.3 of the Public Domain License (PDL) 1.0 states that it is compatible with CC BY 4.0, we have relicensed the dataset under PDL 1.0 to ensure strict consistency with the original licensing terms of the source data.
|
| 137 |
+
|
| 138 |
## Resources
|
| 139 |
- 📃**Paper**: Read our paper for detailed dataset construction pipeline and evaluation results at https://pub.sakana.ai/edinet-bench
|
| 140 |
- 🏗️**Counstruction Code**: Create a new benchmark dataset at https://github.com/SakanaAI/edinet2dataset
|
|
|
|
| 178 |
})
|
| 179 |
```
|
| 180 |
|
|
|
|
| 181 |
|
| 182 |
**Earnings forecast**
|
| 183 |
|
|
|
|
| 227 |
|
| 228 |
## LICENSE
|
| 229 |
|
| 230 |
+
EDINET-Bench is licensed under the [PDL 1.0](https://www.digital.go.jp/resources/open_data/public_data_license_v1.0) in accordance with [EDINET's Terms of Use](https://disclosure2dl.edinet-fsa.go.jp/guide/static/disclosure/WZEK0030.html).
|
| 231 |
|
| 232 |
|
| 233 |
## ⚠️ Warnings
|