Show HN: Spam classifier in Go using Naive Bayes
14 days ago
- #Naive Bayes
- #Spam Classification
- #Go
- Naive Bayes spam classifier implementation in Go
- Uses probabilistic classification based on Bayes' theorem
- Implements Laplace smoothing for handling zero probabilities
- Simple API for training and classifying messages
- Includes tests with real spam/ham email datasets
- Training phase calculates prior probabilities and builds vocabulary
- Classification phase uses log probabilities to avoid underflow
- Supports Kaggle Spam Mails Dataset with an initialization script
- Includes test suite with accuracy measurements