Natural Language Processing Spam Filter

Inspiration

We attempted many project ideas, and this project turned out to be the most accessible considering our knowledge and skills. Besides that, spam filters have typically been made for emails, but it is not widely implemented for text messages. Hence, we aimed to attempt to implement this idea for text messages as well.

What it does

Predicts if an input message is spam or not. User is then able to confirm if the prediction is correct or not, which will then update the model accordingly.

How we built it

Firstly, we obtained the dataset from Kaggle and cleaned the data appropriately. Next, we developed the base model using TFID Vectorizer and SVM. Finally, we saved and implemented the model into a simple UI to demonstrate how it may be applied into an application.

Challenges we ran into

We were new to this field and had to do a lot of research to figure out what everything means and what are the general steps in developing NLP technology.

Accomplishments that we're proud of

We successfully cleaned the data, created the models, and implemented them into a working solution that allows the user to determine if a message is spam or not.

What we learned

We learned the basic concepts of using NLP technology as a spam filter as well the basic ideas of deep learning.

What's next for Natural Language Processing Spam Filter

To implement a better model such as RNN to significantly improve the accuracy. To train the models on a larger dataset as the current dataset is still considered relatively small. To implement the solution into a working mobile application or API

Built With

• python