TikTok Political Sentiment Labeling -- 2
Budget: $250 – $750 USD
I have 5000-6000 TikTok clips that need to be labeled for political sentiment so they can serve as a reliable machine-learning training set. You will watch each video and tag it as either pro-candidate, anti-candidate, or neutral, following a detailed rubric I will supply, you can see the dataset instructions in attached docs. The list of specific candidates we’re tracking is already prepared, and I’ll share it at kickoff along with the exact labeling guidelines and examples.
Because this dataset will feed algorithmic research, accuracy and consistency matter more than speed, yet I do need the bulk of the work finished ASAP. Familiarity with TikTok culture, current political discourse, and subtle forms of persuasion is essential to avoid misclassification.
Scope of work:
• Review 5000 - 6000 short videos in the web-based tool I provide (or your own if you prefer and it exports clean CSV/JSON).
• Apply the sentiment label from the rubric to each clip.
• Flag ambiguous cases so I can adjudicate them quickly.
• Deliver a clean, well-structured export of the labels plus any notes on edge cases.
Ideal background: prior content annotation, sentiment analysis, or dataset preparation; comfort with political topics; ability to stay objective. If you have handled multilingual or meme-heavy content on TikTok, let me know—those skills help.
I’m available for rapid feedback during the project and will run spot-checks to ensure consistency. Let’s get started right away so we can meet the aggressive timeline.
Because this dataset will feed algorithmic research, accuracy and consistency matter more than speed, yet I do need the bulk of the work finished ASAP. Familiarity with TikTok culture, current political discourse, and subtle forms of persuasion is essential to avoid misclassification.
Scope of work:
• Review 5000 - 6000 short videos in the web-based tool I provide (or your own if you prefer and it exports clean CSV/JSON).
• Apply the sentiment label from the rubric to each clip.
• Flag ambiguous cases so I can adjudicate them quickly.
• Deliver a clean, well-structured export of the labels plus any notes on edge cases.
Ideal background: prior content annotation, sentiment analysis, or dataset preparation; comfort with political topics; ability to stay objective. If you have handled multilingual or meme-heavy content on TikTok, let me know—those skills help.
I’m available for rapid feedback during the project and will run spot-checks to ensure consistency. Let’s get started right away so we can meet the aggressive timeline.