The problem Spam_Detection solves

This project helps detect and filter spam messages using machine learning, reducing the risk of scams and unwanted content. It automates text classification to enhance user safety and improve message handling in digital platforms.

Challenges I ran into

One major challenge was minimizing false positives, where normal messages were incorrectly classified as spam. This required careful preprocessing and tuning of the machine learning model. I experimented with different algorithms and finally achieved better accuracy using TF-IDF with Multinomial Naive Bayes.

Another hurdle was integrating the ML model with the Flask API, especially handling serialized files (.pkl). I resolved it by using joblib for model loading and ensuring the text input was properly cleaned before prediction.

Also, managing frontend-backend communication was tricky at first. I fixed CORS issues and used axios in React to send requests smoothly to the Flask server.