PAN23 Profiling Cryptocurrency Influencers with Few-shot Learning

This is the dataset for the shared task on Profiling Cryptocurrency Influencers with Few-shot Learning. Please consult the task's page for further details on the format, the dataset's creation, and links to baselines and utility code. <strong>Task</strong>: In this shared task we aim to profile cryptocurrency influencers in social media, from a low-resource perspective. Moreover, we propose to categorize other related aspects of the influencers, also using a low-resource setting. Specifically, we focus on English Twitter posts for three different sub-tasks: <strong>Low-resource influencer profiling (subtask1):</strong> Input:<br> 32 users per label with a maximum of 10 English tweets each.<br> Classes: (1) null, (2) nano, (3) micro, (4) macro, (5) mega Official evaluation metric: Macro F1 Submission: TIRA. Baselines: User-character Logistic Regression; t5-large (bi-encoders) - zero shot [7], t5-large (label tuning) - few shot [7] <strong>Low-resource influencer interest identification (subtask2):</strong> Input:<br> 64 users per label with 1 English tweet each.<br> Classes: (1) technical information, (2) price update, (3) trading matters, (4) gaming, (5) other Official evaluation metric: Macro F1 Submission: TIRA. Baselines: User-character Logistic Regression; t5-large (bi-encoders) - zero shot [7], t5-large (label tuning) - few shot [7] <strong>Low-resource influencer intent identification (subtask3):</strong> Input:<br> 64 users per label with 1 English tweets each.<br> Classes: (1) subjective opinion, (2) financial information, (3) advertising, (4) announcement Official evaluation metric: Macro F1 Submission: TIRA. Baselines: User-character Logistic Regression; t5-large (bi-encoders) - zero shot [7], t5-large (label tuning) - few shot [7]