---
license: cc-by-3.0
task_categories:
- text-classification
language:
- en
pretty_name: Claim Stance
size_categories:
- 1K<n<10K
configs:
- config_name: claim_stance
  data_files:
  - split: train
    path: "train.csv"
  - split: test
    path: "test.csv"
- config_name: claim_stance_topic
  data_files:
  - split: train
    path: "train_topic.csv"
  - split: validation
    path: "dev_topic.csv"
  - split: test
    path: "test_topic.csv"
---
---

# Dataset Card for Claim Stance Dataset

## Table of Contents

- [Dataset Summary](#dataset-summary)
- [Dataset Structure](#dataset-structure)
- [Licensing Information](#licensing-information)
- [Citation Information](#citation-information)
- [Notes](#notes)

## Dataset Summary 

### Claim Stance

This dataset contains 2,394 labeled Wikipedia claims for 55 topics. The dataset includes the stance (Pro/Con) of each claim towards the topic, 
as well as fine-grained annotations, based on the semantic model of [Stance Classification of Context-Dependent Claims](https://aclanthology.org/E17-1024/) (topic target,
topic sentiment towards its target, claim target, claim sentiment towards its target, and the relation between the targets).  

The dataset is divided into a training set (25 topics, 1,039 claims) and a test set (30 topics, 1,355 claims).

The information in this card refers to this subset of the dataset unless stated otherwise.

### Claim Stance Topic

This subset contains the claims (column `text`) only associated with the topic (column `label`) in a different split to train-validation-test. 
This subset can be utilized for topic classification tasks.

## Dataset Structure

* topicId - internal topic ID
* split - train or test
* topicText - the topic text
* topicTarget - sentiment target of topic
* topicSentiment - topic sentiment towards its target (1:positive/-1:negative)
* claims.claimId - claim internal ID
* claims.stance - PRO or CON
* claims.claimCorrectedText - the corrected version of the claim
* claims.claimOriginalText - the original version of the claim
* claims.Compatible - is the claim compatible with the semantic model of [Stance Classification of Context-Dependent Claims](https://aclanthology.org/E17-1024/)? (yes/no)
        
The following fine-grained annotations are specified only for "compatible" claims
* claims.claimTarget.text - claim sentiment target text (in the corrected version of the claim)
* claims.claimTarget.span.start - 0,
* claims.claimTarget.span.end - 31
* claims.claimSentiment - claim's sentiment towards its target (1:positive/-1:negative)
* claims.targetsRelation - relation between claim target and topic target ((1:consistent/-1:contrastive))

## Licensing Information

The datasets are released under the following licensing and copyright terms:
* (c) Copyright [Wikipedia](https://en.wikipedia.org/wiki/Wikipedia:Copyrights#Reusers.27_rights_and_obligations)
* (c) Copyright IBM 2014. Released under [CC-BY-SA 3.0](http://creativecommons.org/licenses/by-sa/3.0/)

## Citation Information

If you use this dataset, please cite the following paper:

```
@inproceedings{bar-haim-etal-2017-stance,
    title = "Stance Classification of Context-Dependent Claims",
    author = "Bar-Haim, Roy  and
      Bhattacharya, Indrajit  and
      Dinuzzo, Francesco  and
      Saha, Amrita  and
      Slonim, Noam",
    editor = "Lapata, Mirella  and
      Blunsom, Phil  and
      Koller, Alexander",
    booktitle = "Proceedings of the 15th Conference of the {E}uropean Chapter of the Association for Computational Linguistics: Volume 1, Long Papers",
    month = apr,
    year = "2017",
    address = "Valencia, Spain",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/E17-1024",
    pages = "251--261",
    abstract = "Recent work has addressed the problem of detecting relevant claims for a given controversial topic. We introduce the complementary task of Claim Stance Classification, along with the first benchmark dataset for this task. We decompose this problem into: (a) open-domain target identification for topic and claim (b) sentiment classification for each target, and (c) open-domain contrast detection between the topic and the claim targets. Manual annotation of the dataset confirms the applicability and validity of our model. We describe an implementation of our model, focusing on a novel algorithm for contrast detection. Our approach achieves promising results, and is shown to outperform several baselines, which represent the common practice of applying a single, monolithic classifier for stance classification.",
}
```

Improved stance classification results on this dataset were published in:

```
@inproceedings{bar-haim-etal-2017-improving,
    title = "Improving Claim Stance Classification with Lexical Knowledge Expansion and Context Utilization",
    author = "Bar-Haim, Roy  and
      Edelstein, Lilach  and
      Jochim, Charles  and
      Slonim, Noam",
    editor = "Habernal, Ivan  and
      Gurevych, Iryna  and
      Ashley, Kevin  and
      Cardie, Claire  and
      Green, Nancy  and
      Litman, Diane  and
      Petasis, Georgios  and
      Reed, Chris  and
      Slonim, Noam  and
      Walker, Vern",
    booktitle = "Proceedings of the 4th Workshop on Argument Mining",
    month = sep,
    year = "2017",
    address = "Copenhagen, Denmark",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/W17-5104",
    doi = "10.18653/v1/W17-5104",
    pages = "32--38",
    abstract = "Stance classification is a core component in on-demand argument construction pipelines. Previous work on claim stance classification relied on background knowledge such as manually-composed sentiment lexicons. We show that both accuracy and coverage can be significantly improved through automatic expansion of the initial lexicon. We also developed a set of contextual features that further improves the state-of-the-art for this task.",
}
```

## Notes

(1) Claim annotations and the experiments reported in [Stance Classification of Context-Dependent Claims](https://aclanthology.org/E17-1024/) and [Improving Claim Stance Classification with Lexical Knowledge Expansion and Context Utilization](https://aclanthology.org/W17-5104/)
    are based on the corrected version of the claim. See [A Benchmark Dataset for Automatic Detection of Claims and Evidence in the Context of Controversial Topics](https://aclanthology.org/W14-2109/) for description of generating 
    corrected version for claims. The original version is the claim as it is found in the clean version of 
    the article, with no further editing.

(2) The topics and claims partially overlap with the CE-EMNLP-2015 dataset:
    Common topics IDs: 1, 21, 61, 81, 101, 121, 181, 221, 323, 381, 441, 442, 443, 481, 482, 483, 601, 602, 
    621, 641, 642, 644, 645, 648, 662, 663, 665, 681, 683, 701, 721, 742, 743, 744, 761, 801, 803, 841, 861, 
    881, 923, 926, 941, 942, 944, 946
    Only this dataset: 603, 661, 922, 985, 987, 990, 994, 1005, 1065
    Only the CE-EMNLP-2015 dataset: 643, 646, 647, 664, 821, 902, 921, 925, 943, 945, 947, 961