Academic Stance on AI in The Conversation Articles, Analyzed via Keyness

A text corpus of academic articles from The Conversation discussing artificial intelligence, analyzed for stance markers. The 4.7 MB dataset was created by Margo Van Poucke and last updated on 2026-04-25. It supports a study using Kullback-Leibler Divergence and dispersion measures to identify lexicogrammatical patterns of stance.

Use Cases

• Analyzing the use of epistemic markers (e.g., 'likely', 'could') in expert public communication about AI.
• Studying attitudinal resources (e.g., 'useful', 'difficult') to understand the framing of AI's opportunities and ethical dilemmas.
• Comparing stance marker density between academic news and general news registers to investigate hybrid discourse.
• Training models to identify cautiously optimistic language in technology-related public discourse.

Strengths

• Dataset size is 4.7 MB, indicating a substantive text corpus.
• Analysis employs specific, documented methodologies: Kullback-Leibler Divergence (KLD) and Gries’ DPnorm.
• Source is clearly defined as articles from The Conversation, a specific publication platform.

Limitations

• Row count and specific column/feature structure are unknown, limiting suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Freshness should be verified as the last update date is 2026-04-25.

Provenance

Source

Articles published in The Conversation.

Collection Method

Semi-supervised automated retrieval and syntactic parsing with SpaCy, followed by keyness analysis.

Freshness

Last updated 2026-04-25 00:10:09.