This project focuses on detecting emotional sentiment in Vietnamese text using a rule-based approach. It involves building a custom emotion lexicon and applying logical rules to classify input sentences into emotion categories such as positive or negative.
- Source: Manually created dataset of Vietnamese sentences
- Format:
.txtand.csvfiles for sample text and emotion word dictionary - Categories: Positive (1), Negative (-1)
- Vietnamese text preprocessing: lowercasing, punctuation removal, tokenization
- Dictionary-based sentiment scoring using positive/negative word lists
- Rule-based classification logic (count-based comparison)
- Accuracy evaluation by comparing predicted vs. labeled sentiments
- Batch sentiment prediction from tokenized CSV files
- Build a transparent, interpretable rule-based emotion detection system for Vietnamese
- Perform basic sentiment classification without using machine learning
- Provide a lightweight prototype for future improvement or extension
- The predicted labels (
pre_Sentiment) matched ground-truth labels (Sentiment) on ~84.7% of test samples - Results are saved to CSV for easy inspection, validation, or dashboard integration
- Python, Pandas, re (Regex)
- Text preprocessing techniques: lowercasing, punctuation removal, tokenization
⭐ This project was developed during my learning journey and reflects my ability to apply concepts into practice. It continues to be improved as I grow