---
dataset_info:
  features:
  - name: sent_id
    dtype: string
  - name: tokens
    list: string
  - name: head
    list: string
  - name: deprel
    list:
      class_label:
        names:
          '0': acl
          '1': acl:relcl
          '2': advcl
          '3': advcl:cleft
          '4': advmod
          '5': amod
          '6': appos
          '7': aux
          '8': aux:caus
          '9': aux:pass
          '10': aux:tense
          '11': case
          '12': cc
          '13': ccomp
          '14': compound
          '15': conj
          '16': cop
          '17': csubj
          '18': csubj:pass
          '19': dep
          '20': dep:comp
          '21': det
          '22': discourse
          '23': dislocated
          '24': expl:comp
          '25': expl:pass
          '26': expl:subj
          '27': fixed
          '28': flat:foreign
          '29': flat:name
          '30': iobj
          '31': mark
          '32': nmod
          '33': nsubj
          '34': nsubj:caus
          '35': nsubj:outer
          '36': nsubj:pass
          '37': nummod
          '38': obj
          '39': obj:lvc
          '40': obl
          '41': obl:agent
          '42': obl:arg
          '43': obl:mod
          '44': orphan
          '45': parataxis
          '46': parataxis:insert
          '47': punct
          '48': reparandum
          '49': root
          '50': vocative
          '51': xcomp
  splits:
  - name: train
    num_bytes: 9916993
    num_examples: 17925
  - name: dev
    num_bytes: 1361562
    num_examples: 2950
  - name: test_fr_gsd
    num_bytes: 231864
    num_examples: 415
  - name: test_fr_rhapsodie
    num_bytes: 261360
    num_examples: 812
  - name: test_fr_sequoia
    num_bytes: 239070
    num_examples: 456
  download_size: 2551975
  dataset_size: 12010849
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: dev
    path: data/dev-*
  - split: test_fr_gsd
    path: data/test_fr_gsd-*
  - split: test_fr_rhapsodie
    path: data/test_fr_rhapsodie-*
  - split: test_fr_sequoia
    path: data/test_fr_sequoia-*
---

# Dataset Information

This dataset is designed for the **Dependency parsing task** and is used to train the Airudit multitask model.

## Dataset Description

The dataset combines **Universal Dependencies v2.17** French corpora available at  
[commul/universal_dependencies](https://huggingface.co/datasets/commul/universal_dependencies).

Included corpora:

- commul/universal_dependencies/fr_gsd
- commul/universal_dependencies/fr_sequoia
- commul/universal_dependencies/fr_rhapsodie

Nb. Other corpora, as 'fr_parisstories', 'fr_partut', 'fr_poitevindivital', 'fr_pud' are excluded because of too large difference in labels sets.

## Dataset Structure

The dataset contains the following splits:

```
DatasetDict({
    train: Dataset({
        features: ['sent_id', 'tokens', 'head', 'deprel'],
        num_rows: 17925
    })
    dev: Dataset({
        features: ['sent_id', 'tokens', 'head', 'deprel'],
        num_rows: 2950
    })
    test_fr_gsd: Dataset({
        features: ['sent_id', 'tokens', 'head', 'deprel'],
        num_rows: 415
    })
    test_fr_rhapsodie: Dataset({
        features: ['sent_id', 'tokens', 'head', 'deprel'],
        num_rows: 812
    })
    test_fr_sequoia: Dataset({
        features: ['sent_id', 'tokens', 'head', 'deprel'],
        num_rows: 456
    })
})
```

## Labels
Labels are encoded in the dataset and can be retrieved as following : 

```python
from datasets import load_dataset
ds = load_dataset('airudit/UD_v2_17_DEP')
ds["train"].features["deprel"].feature.names
# ['acl', 'acl:relcl', 'advcl', 'advcl:cleft', 'advmod', 'amod', 'appos', 'aux', 'aux:caus', 'aux:pass', 'aux:tense', 'case', 'cc', 'ccomp', 'compound', 'conj', 'cop', 'csubj', 'csubj:pass', 'dep', 'dep:comp', 'det', 'discourse', 'dislocated', 'expl:comp', 'expl:pass', 'expl:subj', 'fixed', 'flat:foreign', 'flat:name', 'iobj', 'mark', 'nmod', 'nsubj', 'nsubj:caus', 'nsubj:outer', 'nsubj:pass', 'nummod', 'obj', 'obj:lvc', 'obl', 'obl:agent', 'obl:arg', 'obl:mod', 'orphan', 'parataxis', 'parataxis:insert', 'punct', 'reparandum', 'root', 'vocative', 'xcomp']
```

## 🔍 Quick usage example: 

```python
from datasets import load_dataset
ds = load_dataset('airudit/UD_v2_17_DEP')
ds["train"][0]
# {'sent_id': [...], 'tokens': [...], 'head': [...], 'deprel': [...]}
```

## 📦 Dataset Pre-Processing  :

### Dependency Relation Filtering : 
Sentences containing the following dependency relations are removed:
- [flat](https://universaldependencies.org/treebanks/fr_gsd/fr_gsd-dep-flat.html) This relation does not appear in **Sequoia** treebank beacause the annotation always uses a subtype (e.g. `flat:name`). 
   When `flat` is not subtyped it is difficult to automatically map it to a consistent subype.
- [parataxis:parenth](https://universaldependencies.org/fr/dep/parataxis-parenth.html) is used to annotate parenthetical clauses.
  Only present in **Rhapsodie** treebank.
  The number of occurrences is small (39 occ), and parentical are currently not cosidered essential for target multitask model.

In total, removed **14** sentences for **GSD** treebank and **78** sentences for **Rhapsodie** treebank.

### Dependency Relation Normalization : 

To harmonize the annotation schemes across treebanks, the following mapping is applied:
```
{
  "goeswith": "compound",
  "expl:pv": "expl:pass",
  "obj:agent": "obj",
  "iobj:agent": "iobj",
  "nmod:appos": "nmod"
}
```

Occurrences affected by this normalization:
```
fr_gsd
{'expl:pv': 1016, 'obj:agent': 111, 'goeswith': 38, 'iobj:agent': 24}
fr_rhapsodie
{'nmod:appos': 115}
fr_sequoia
{'expl:pv': 242, 'obj:agent': 12, 'goeswith': 2, 'iobj:agent': 1}
```

### Additional Label

The relation [reparandum](https://universaldependencies.org/treebanks/fr_rhapsodie/fr_rhapsodie-dep-reparandum.html) is added to the final vocabulary.
- The label appears only in **Rhapsodie** treebank (1205 occ).
- It is used to annotate speech disfluencies and self-corrections.
- This information can be useful for ASR post-processing tasks.

### Structure
- Train and dev splits from all corpora are merged into unified train and dev splits.
- Test splits are preserved and renamed according to their source corpus.
- Only the relevant columns are kept: ["sent_id", "tokens", "head", "deprel"].
- The dataset was generated using the following script: multitask-nlp/src/multitask_nlp/datasets/dependency_dataset_preparation.py