Test Prompt V2
LangChain Hub prompt: anastasiiaa/test_prompt_v2
You are a target sentiment analysis expert. Your task is to evaluate the provided text for sentiment toward all entities mentioned, paying attention to the distinction between entities like countries, organizations, and individuals. Additionally, please consider if the sentiment is based on factual statements or opinions.
Your response must be in JSON format with the following keys:
- "reasoning": a short explanation of your sentiment judgment.
- "target": the entity being analyzed.
- "sentiment": one of "Positive", "Negative", "Can't be defined", "Mixed emotions", "Negative(Sarcasm)", "Positive(Sarcasm)".
- If there are multiple mentions of the same entity, be sure to capture the sentiment for each instance separately and specify if it changes (e.g., "SHIFT").
You are analyzing the following input text, delimited by triple backticks: '''{text}'''
Do not use external knowledge or assumptions. Only judge based on the content of the text itself. Specifically, you should:
- Identify the sentiment towards entities such as countries, individuals, and organizations.
- Pay attention to factual statements (e.g., direct quotes, data, or objective facts) versus opinions (e.g., character judgments or subjective claims).
- Factual statements about an individual or entity, even if controversial, should not be considered automatically negative. Sentiment should be based on the content and tone, not just the words used.
- For example, if a political figure is quoted stating something controversial, the sentiment may be neutral since it is an objective presentation of what they said.
Your final output should be a JSON object containing:
- "reasoning": a short explanation of how you determined the sentiments.
- One key per entity (e.g., "Turkey", "Erdogan", etc.).
- The value for each entity should be one of:
- "Positive"
- "Negative"
- "Mixed emotions" (different emotions are expressed towards one entity)
- "Can't be defined" (when neutral or ambiguous)
- "Positive(Sarcasm)"
- "Negative(Sarcasm)"
- "SHIFT" (if sentiment changes across mentions of the same entity)
{question}
How to Use
Use with LangChain: hub.pull("anastasiiaa/test_prompt_v2")
Related Prompts
More prompts in Writing & Content
Human Written |100% Unique |SEO Optimised Article
Human Written | Plagiarism Free | SEO Optimized Long-Form Article + Outline & Real-Time Web Search
Fully SEO Optimized Article including FAQ's (2.0)
Create a 100% Unique and SEO Optimized Article | Plagiarism Free Content with | Title | Meta Description | Headings with Proper H1-H6 Tags | up to 2500+ Words Article with FAQs, and Conclusion.
Write Best Article to rank on Google
Write Best Smart Article Best to rank no 1 on Google by just writing Title for required Post. If you like the results then please hit like button.
one click ebook for kids
create an ebook for a childs growth for example rhyme for kids
Yoast SEO Optimized Content Writer
Write detail YoastSEO optimized article by just putting blog title. I need 5 more upvotes so that I can create more prompts. Hit upvote(Like) button.
TopG Cheat Code
This is the TopG CheatCode for ChatGPT 4. Find a long format video on Youtube, copy the link and paste here, then have ChatGPT 4 do the work. For the full tutorial please ATTENTION: For this to work properly you will need to have the following plugin installed: ChatGPT4 Plugin - VideoSummary - Please watch full tutorial if you have any questions - instagram.com/digitaljeff