strombergnlp/nlpcc-stance

Dataset Overview

Basic Information

• Name: NLPCC 2016: Stance Detection in Chinese Microblogs
• Language: Chinese (bcp47:zh)
• License: Creative Commons Attribution 4.0 (CC‑BY‑4.0)
• Multilinguality: Monolingual
• Size: 1K < n < 10K
• Source: Original data
• Task Category: Text classification
• Task ID: Sentiment analysis
• Tags: Stance detection

Description

• Overview: This dataset focuses on stance prediction in Chinese micro‑blogs. It originates from the NLPCC‑ICCPOL 2016 shared task, a mandatory supervised task that detects stance toward five interest targets.
• Supported Task: Stance detection in Chinese micro‑blogs

Structure

• Instances: Each instance includes four fields: id (unique identifier), target (stance target), text (text containing the stance), stance (stance class, 0 = AGAINST, 1 = FAVOR, 2 = NONE).
• Fields:
  • id: String, unique identifier
  • target: String, stance target
  • text: String, text containing the stance
  • stance: Integer, stance class (0: AGAINST, 1: FAVOR, 2: NONE)
• Splits: Training set contains 2,986 instances

Creation

• Motivation: To create a stance‑annotated dataset for micro‑blog texts. Six stance targets were selected and data were collected from Sina Weibo for annotation.
• Source Data: Content generated by Sina Weibo users
• Annotation:
  • Each target‑post pair was independently labeled by two students. If they agreed, the label was accepted; otherwise a third annotator resolved the disagreement through voting.
  • Annotators: Chinese students

Considerations

• Social Impact: The dataset retains the original social‑media statements, which may involve privacy concerns.
• Bias Discussion: The data exhibit temporal, geographic, and topical biases.

Additional Information

• Dataset Curator: Paper authors
• License Information: Distributed under CC‑BY‑4.0
• Citation:

  @incollection{xu2016overview,
    title={Overview of nlpcc shared task 4: Stance detection in chinese microblogs},
    author={Xu, Ruifeng and Zhou, Yu and Wu, Dongyin and Gui, Lin and Du, Jiachen and Xue, Yun},
    booktitle={Natural language understanding and intelligent applications},
    pages={907--916},
    year={2016},
    publisher={Springer}
  }