This projects use native bayes algorithm to detect spam content, like normal, sex, gamble, politics content. We use 400 Chinese sentences to train the algorithm and the correct rate is about 93%
- Fast We offer built-in cache files so it's super fast even when you first used
- High accuracy Our correct rate is about 93%
- Customizable You can add what data you need to train