Detect text sentiment: positive, neutral or negative.
Prompt Content
Specify the sentiment of the following titles, assigning them the values of: positive, neutral or negative. Generate the results in column, including the titles in the first one, and their sentiment in the second: [text batch to analyze]
FAQ
What if the sentiment classification is wrong?
The AI often misreads sarcasm or passive aggression. For this kind of text, append "annotate both literal sentiment and actual sentiment (accounting for sarcasm), and explain the reasoning." Dual annotation exposes the AI's misjudgments and makes manual correction much easier.
Can I batch analyze 100 comments?
Yes, but watch the context limit. Keep it under 30-50 per run. With too much input the AI gets lazy and labels the second half with one sentiment. Run in batches and require "output one line per input in original order, do not merge."
How do I use this prompt?
Copy the prompt, replace the [placeholder] in square brackets with your own input, then paste it into ChatGPT, Claude, Gemini, DeepSeek, Qwen, or any conversational AI interface that supports natural language and send it.