Skip to content

Latest commit

 

History

History
17 lines (11 loc) · 805 Bytes

statistics.md

File metadata and controls

17 lines (11 loc) · 805 Bytes

Statistics of TwiBot-22

TwiBot-22 is the largest and most comprehensive Twitter bot detection benchmark to date. Specifically, TwiBot-20 is designed to address the challenges of limited dataset scale, imcomplete graph structure, and low annotation quality in previous datasets. For more details, please refer to the TwiBot-22 paper and statistics.

Entity types

entity_type

Relation types

relation_type

Detailed statistics

detailed_sta

Comparing to other datasets

compare

For more detailed information on sampling methods, labeling quality, etc., please refer to TwiBot-22 paper