70,427 Cross-Linked Twitter-GitHub User Pairs for Developer Communication Research

70,427 user pairs linking Twitter and GHTorrent accounts, identified as likely belonging to the same open-source developers. The dataset accompanies a 2020 research paper from Carnegie Mellon University authors Hongbo Fang, Daniel Klug, Hemank Lamba, James Herbsleb, and Bogdan Vasilescu. It is intended for research on how developers discuss their GitHub work on social media.

Use Cases

• Study cross-platform developer identity and self-presentation based on linked Twitter and GitHub accounts.
• Analyze communication patterns about software development work based on the social media linkage.
• Train or evaluate user identity resolution algorithms based on the provided user ID pairs.
• Investigate the relationship between GitHub activity and social media discourse based on the user linkage.

Strengths

• Contains 70,427 user pairs, providing a substantial sample for analysis.
• Links two major platforms (Twitter and GitHub) relevant to developer studies.
• Accompanies a peer-reviewed research paper from a recognized academic institution.

Limitations

• Row count for the underlying user data from Twitter and GitHub is unknown; scale of analysis is user-dependent.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

Carnegie Mellon University research team, via the paperswithcode platform.

Collection Method

User pairs were identified as likely belonging to the same users, linking Twitter and GHTorrent (GitHub) data.