Messages either received in the form of Emails or SMS, sometimes contain useful information or contain some information which is irrelevant, such as Advertisements, or information that may be used to dupe the recipient.
Messages containing useful information are known as HAM Messages, whereas messages containing irrelevant information are known as SPAM Messages.
The project aims to develop ML Models, that can be used detect whether a particular message is a SPAM or HAM Message.
Two approaches were tried to detect messages,
1️⃣ Bag Of Words Model
2️⃣ Word Embeddings
It was found that the Word Embeddings Model has a higher Accuracy & F1 Score.
The following libraries are used for the implementation of the project:
1️⃣ Tensorflow
2️⃣ Keras
3️⃣ Numpy
4️⃣ Pandas
5️⃣ Plotly
6️⃣ Matplotlib
7️⃣ Pickle
8️⃣ h5py
9️⃣ Scikit-Learn