Twitter Misinformation Dataset for Training Detection Models

A compilation of several existing datasets focused on misinformation detection, disaster-related tweets, and fact-checking. The dataset was created by author roupenminassian and last updated on November 24, 2024. It has been utilized in research studying backdoor attacks in textual content.

Use Cases

• Train misinformation detection models based on the combined data from multiple fact-checking sources.
• Study disaster-related misinformation patterns based on the included disaster-related tweets.
• Research backdoor attacks in textual content, as referenced in the description's mention of 'Claim-Guided Textual...'.

Strengths

• Combines data from multiple existing sources, which likely increases topic diversity.
• Has been used in published research on backdoor attacks, indicating a level of academic validation.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and license information are unknown, which may limit suitability assessment.
• Last updated 2024-11-24 06:10:20; freshness should be verified.

Provenance

Source

Multiple existing datasets on misinformation detection and disaster-related tweets, compiled by roupenminassian.

Collection Method

Compilation of existing datasets.

Freshness

Last updated 2024-11-24 06:10:20.