metadata

dataset_info:
  features:
    - name: id_doc
      dtype: string
    - name: words
      sequence:
        sequence:
          sequence: string
    - name: lemmas
      sequence:
        sequence:
          sequence: string
    - name: msds
      sequence:
        sequence:
          sequence: string
    - name: ne_tags
      sequence:
        sequence:
          sequence: string
    - name: mentions
      list:
        - name: id_mention
          dtype: string
        - name: mention_data
          struct:
            - name: idx_par
              dtype: uint32
            - name: idx_sent
              dtype: uint32
            - name: word_indices
              sequence: uint32
            - name: global_word_indices
              sequence: uint32
    - name: coref_clusters
      sequence:
        sequence: string
  splits:
    - name: train
      num_bytes: 21547216
      num_examples: 756
  download_size: 21892324
  dataset_size: 21547216
license: cc-by-sa-4.0
language:
  - sl
pretty_name: SentiCoref
size_categories:
  - n<1K

Dataset card for SentiCoref

Usage

import datasets
data = datasets.load_dataset("cjvt/senticoref", trust_remote_code=True)

Dataset Summary

The dataset contains the SentiCoref corpus, annotated for coreference. It is part of the SUK training bundle of corpora. For more details please check the paper or the Clarin repository from which this dataset is being loaded.

Dataset Structure

Data Instances

{
    'id_doc': 'senticoref1', 
    'words': [
        [
            ['Evropska', 'komisija', 'mora', 'narediti', 'analizo', 'vzrokov', 'rasti', 'cen', 'hrane', ',', 'menita', 'kmetijski', 'minister', 'Jarc', 'in', 'njegov', 'francoski', 'kolega', '.'], 
            ['Bo', 'evropska', 'komisija', 'analizirala', 'vzroke', 'rasti', 'cen', 'hrane', '.'], 
            ...
        ],
        ...
    ], 
    'lemmas': [
        [
            ['evropski', 'komisija', 'morati', 'narediti', 'analiza', 'vzrok', 'rast', 'cena', 'hrana', ',', 'meniti', 'kmetijski', 'minister', 'Jarc', 'in', 'njegov', 'francoski', 'kolega', '.'], 
            ['biti', 'evropski', 'komisija', 'analizirati', 'vzrok', 'rast', 'cena', 'hrana', '.'],
            ...
        ]
    ], 
    'msds': [
        [
            ['mte:Ppnzei', 'mte:Sozei', 'mte:Ggnste', 'mte:Ggdn', 'mte:Sozet', 'mte:Sommr', 'mte:Sozer', 'mte:Sozmr', 'mte:Sozer', 'mte:U', 'mte:Ggnstd', 'mte:Ppnmeid', 'mte:Somei', 'mte:Slmei', 'mte:Vp', 'mte:Zstmeiem', 'mte:Ppnmeid', 'mte:Somei', 'mte:U'], 
            ['mte:Gp-pte-n', 'mte:Ppnzei', 'mte:Sozei', 'mte:Ggvd-ez', 'mte:Sommt', 'mte:Sozer', 'mte:Sozmr', 'mte:Sozer', 'mte:U'], 
            ...
        ], 
        ...
    ],
    'ne_tags': [
        [
            ['B-ORG', 'I-ORG', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B-PER', 'O', 'O', 'O', 'O', 'O'], 
            ['O', 'B-ORG', 'I-ORG', 'O', 'O', 'O', 'O', 'O', 'O'], 
            ...
        ], 
        ...
    ], 
    'mentions': [
        {'id_mention': 'senticoref1.1.1.ne1', 'mention_data': {'idx_par': 0, 'idx_sent': 0, 'word_indices': [0, 1], 'global_word_indices': [0, 1]}}, 
        ...
    ], 
    'coref_clusters': [
        ['senticoref1.1.1.ne1', 'senticoref1.1.2.ne1', 'senticoref1.1.3.ne1'], 
        ['senticoref1.1.1.phr52-1', 'senticoref1.1.3.phr52-2', 'senticoref1.1.11.phr52-3'], 
        ['senticoref1.1.1.t5', 'senticoref1.1.3.t6', 'senticoref1.1.11.t11', 'senticoref1.1.11.t17'], 
        ['senticoref1.1.1.phr13-1', 'senticoref1.1.2.phr13-2'], 
        ...
    ]
}

Data Fields

id_doc: a string ID of the document (corresponds to file name in this case);
words: a List[List[List[String]]] containing document words;
lemmas: a List[List[List[String]]] containing document lemmas;
msds: a List[List[List[String]]] containing document morphosyntactic features, encoded using MULTEXT-East V6;
ne_tags: a List[List[List[String]]] containing document named entity tags, encoded using IOB2 scheme;
mentions: a list of dicts for each mention. Each mention contains an ID (id_mention) and positions of words inside mention (determined by idx_sent, word_indices; or equivalently global_word_indices if sentences are flattened into a single list)
coref_clusters: a list of lists of strings containing mention IDs contained inside each coreference cluster.

Additional Information

Dataset Curators

Špela Arhar Holdt; et al. (please see http://hdl.handle.net/11356/1959 for the full list of contributors)

Licensing Information

CC BY-SA 4.0

Citation Information

@article{senticoref-paper,
  title={Neural coreference resolution for Slovene language},
  author={Matej Klemen and Slavko Žitnik},
  journal={Computer Science and Information Systems},
  year={2022},
  volume={19},
  pages={495-521}
}

@misc{suk-clarin,
    title = {Training corpus {SUK} 1.1},
    author = {Arhar Holdt, {\v S}pela and Krek, Simon and Dobrovoljc, Kaja and Erjavec, Toma{\v z} and Gantar, Polona and {\v C}ibej, Jaka and Pori, Eva and Ter{\v c}on, Luka and Munda, Tina and {\v Z}itnik, Slavko and Robida, Nejc and Blagus, Neli and Mo{\v z}e, Sara and Ledinek, Nina and Holz, Nanika and Zupan, Katja and Kuzman, Taja and Kav{\v c}i{\v c}, Teja and {\v S}krjanec, Iza and Marko, Dafne and Jezer{\v s}ek, Lucija and Zajc, Anja},
    url = {http://hdl.handle.net/11356/1959},
    note = {Slovenian language resource repository {CLARIN}.{SI}},
    copyright = {Creative Commons - Attribution-{ShareAlike} 4.0 International ({CC} {BY}-{SA} 4.0)},
    issn = {2820-4042},
    year = {2024}
}

Contributions

Thanks to @matejklemen for adding this dataset.