---
language:
- it
language_details: it-IT
license: cc-by-nc-sa-4.0
task_categories:
- text-classification
task_ids:
- named-entity-recognition
size_categories:
- 10K<n<100K
configs:
- config_name: adg
  data_files:
  - split: train
    path: "adg_train.jsonl"
  - split: dev
    path: "adg_dev.jsonl"
  - split: test
    path: "adg_test.jsonl"
  - split: trial
    path: "adg_trial.jsonl"
  - split: reduced_test
    path: "sampled_adg_test.jsonl"
- config_name: fic
  data_files:
  - split: train
    path: "fic_train.jsonl"
  - split: dev
    path: "fic_dev.jsonl"
  - split: test
    path: "fic_test.jsonl"
  - split: trial
    path: "fic_trial.jsonl"
  - split: reduced_test
    path: "sampled_fic_test.jsonl"
- config_name: wn
  data_files:
  - split: train
    path: "wn_train.jsonl"
  - split: dev
    path: "wn_dev.jsonl"
  - split: test
    path: "wn_test.jsonl"
  - split: trial
    path: "wn_trial.jsonl"
  - split: reduced_test
    path: "sampled_wn_test.jsonl"
---

# Data for the NERMuD shared task (Evalita 2023)

This data is the one used for the [NERMuD shared task](https://nermud.fbk.eu/) organized
at [Evalita 2023](https://www.evalita.it/campaigns/evalita-2023/).

The dataset contains the Wikinews, fiction, and De Gasperi subsets of [KIND](https://github.com/dhfbk/KIND), where test data is used for development.

## Content of the dataset

| Split                       |    Sentences |
|:----------------------------|-------------:|
| wn_train                    |       10,912 |
| wn_dev                      |        2,594 |
| wn_test                     |        2,088 |
| fic_train                   |       11,423 |
| fic_dev                     |        1,051 |
| fic_test                    |        1,517 |
| adg_train                   |        5,147 |
| adg_dev                     |        1,122 |
| adg_test                    |          521 |

| Set                         |    Sentences |
|:----------------------------|-------------:|
| Total train                 |       27,482 |
| Total dev                   |        4,767 |
| Total test                  |        4,126 |
| Total                       |       36,380 |


## Publications 

```
@inproceedings{evalita2023nermud,
    title={{NERMuD} at {EVALITA} 2023: Overview of the Named-Entities Recognition on Multi-Domain Documents Task},
    author={Palmero Aprosio, Alessio and Paccosi, Teresa},
    booktitle={Proceedings of the Eighth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2023)},
    publisher = {CEUR.org},
    year = {2023},
    month = {September},
    address = {Parma, Italy}
}
```