Customer Reviews Sentiment Analysis
Budget: ₹750 – ₹1,250 INR
I have a folder full of plain .txt files, each containing a customer review from our store. I need those reviews classified for sentiment so I can see, at a glance, how people really feel about us.
Here’s what I’m after:
• Read every text file, clean the content, and run a reliable sentiment-analysis model (Python + spaCy, NLTK, TextBlob, or Hugging Face transformers are all fine—use whatever gets the best accuracy).
• Label each review Positive, Neutral, or Negative and include a confidence score.
• Return a single CSV that lists: file name, full review text, sentiment label, and confidence.
• Provide a brief notebook or Markdown report that explains your steps and visualizes overall sentiment distribution.
Acceptance criteria
• Scripts run locally in one command with requirements.txt.
• Model reaches at least 0.80 macro-F1 on a 10 % hold-out sample you create from the data.
If you’ve tackled customer-review sentiment before, let me know—I’ll be checking portfolios.
Here’s what I’m after:
• Read every text file, clean the content, and run a reliable sentiment-analysis model (Python + spaCy, NLTK, TextBlob, or Hugging Face transformers are all fine—use whatever gets the best accuracy).
• Label each review Positive, Neutral, or Negative and include a confidence score.
• Return a single CSV that lists: file name, full review text, sentiment label, and confidence.
• Provide a brief notebook or Markdown report that explains your steps and visualizes overall sentiment distribution.
Acceptance criteria
• Scripts run locally in one command with requirements.txt.
• Model reaches at least 0.80 macro-F1 on a 10 % hold-out sample you create from the data.
If you’ve tackled customer-review sentiment before, let me know—I’ll be checking portfolios.