SCITE / README.md
thagen's picture
updated feature description
86f195e
metadata
license: cc-by-nc-4.0
task_categories:
  - text-classification
  - token-classification
language:
  - en
multilinguality:
  - monolingual
size_categories:
  - 1K<n<10K
tags:
  - causality
pretty_name: SCITE
paperswithcode_id: ../paper/causality-extraction-based-on-self-attentive
configs:
  - config_name: causality detection
    data_files:
      - split: train
        path: causality-detection/train.parquet
      - split: test
        path: causality-detection/test.parquet
    features:
      - name: index
        dtype: string
      - name: text
        dtype: string
      - name: label
        dtype:
          class_label:
            names:
              '0': uncausal
              '1': causal
  - config_name: causal candidate extraction
    data_files:
      - split: train
        path: causal-candidate-extraction/train.parquet
      - split: test
        path: causal-candidate-extraction/test.parquet
    features:
      - name: index
        dtype: string
      - name: tokens
        sequence: string
      - name: entity
        sequence:
          sequence: int32
  - config_name: causality identification
    data_files:
      - split: train
        path: causality-identification/train.parquet
      - split: test
        path: causality-identification/test.parquet
    features:
      - name: index
        dtype: string
      - name: text
        dtype: string
      - name: relations
        list:
          - name: relationship
            dtype:
              class_label:
                names:
                  '0': no-rel
                  '1': causal
          - name: first
            dtype: string
          - name: second
            dtype: string
train-eval-index:
  - config: causality detection
    task: text-classification
    task_id: text_classification
    splits:
      train_split: train
      eval_split: test
    col_mapping:
      text: text
      label: label
    metrics:
      - type: accuracy
      - type: precision
      - type: recall
      - type: f1
  - config: causal candidate extraction
    task: token-classification
    task_id: token_classification
    splits:
      train_split: train
      eval_split: test
    metrics:
      - type: accuracy
      - type: precision
      - type: recall
      - type: f1
  - config: causality identification
    task: text-classification
    task_id: text_classification
    splits:
      train_split: train
      eval_split: test
    metrics:
      - type: accuracy
      - type: precision
      - type: recall
      - type: f1

This repository integrates the SCITE extended SemEval 2010 Task 8 dataset into hf datasets. It is in conformance with SCITE's CC BY-NC 4.0 license. Please find the original dataset here. Please see the citations at the end of this README.

Dataset Description

Usage

Causality Detection

from datasets import load_dataset
dataset = load_dataset("webis/SCITE", "causality detection")

Causal Candidate Extraction

from datasets import load_dataset
dataset = load_dataset("webis/SCITE", "causal candidate extraction")

Causality Identification

from datasets import load_dataset
dataset = load_dataset("webis/SCITE", "causality identification")

Citations

The SCITE paper by Li et al., 2021:

@article{li:2021,
  title = {Causality Extraction Based on Self-Attentive {{BiLSTM-CRF}} with Transferred Embeddings},
  author = {Li, Zhaoning and Li, Qi and Zou, Xiaotian and Ren, Jiangtao},
  year = {2021},
  journal = {Neurocomputing},
  volume = {423},
  pages = {207--219},
  doi = {10.1016/J.NEUCOM.2020.08.078}
}

SemEval 2010 Task 8 by Hendrickx et al., 2010 which SCITE builds upon:

@inproceedings{hendrickx:2010,
  title = {{{SemEval-2010 Task}} 8: {{Multi-Way Classification}} of {{Semantic Relations}} between {{Pairs}} of {{Nominals}}},
  shorttitle = {{{SemEval-2010 Task}} 8},
  booktitle = {Proceedings of the 5th {{International Workshop}} on {{Semantic Evaluation}}, {{SemEval}}@{{ACL}} 2010, {{Uppsala University}}, {{Uppsala}}, {{Sweden}}, {{July}} 15-16, 2010},
  author = {Hendrickx, Iris and Kim, Su Nam and Kozareva, Zornitsa and Nakov, Preslav and S{\'e}aghdha, Diarmuid {\'O} and Pad{\'o}, Sebastian and Pennacchiotti, Marco and Romano, Lorenza and Szpakowicz, Stan},
  editor = {Erk, Katrin and Strapparava, Carlo},
  year = {2010},
  pages = {33--38},
  publisher = {The Association for Computer Linguistics}
}