Spam filter for Quora questions

This is project no. 3

In this project the task is to filter the questions asked in quora to see if they are spam or not.

The projcct is a binary classification as such:

• spam:0
• not spam:1

Before we applying any deep learning algorith to it we first need to process each question to clean and remove some of the following things as such:

• convert the entire sentences to lower case
• remove any punctuation
• white space removal
• remove any new line
• remove number etc,

Now the process is to create features out of word, we accomplished it using (glove embedding)[http://nlp.stanford.edu/data/glove.42B.300d.zip].

embeding_index={}

f=open('glove.42B.300d.txt',encoding='utf-8')

for line in f:
   values=line.split()
   word=values[0]
   coefs=np.asarray(values[1:],dtype='float32')
   embeding_index[word]=coefs
f.close()

And using the same embedding we are going to create weights for our network:

embedding_matrix=np.zeros((vocab_size+1,300))

for word,i in tk.word_index.items():
   embed_vector=embeding_index.get(word)
   if embed_vector is not None:
       embedding_matrix[i]=embed_vector

Finally we will use these in the network as such:

inputs=Input(name='text_input',shape=[max_len,])

embed=Embedding(vocab_size+1,
               300,
               input_length=max_len,
               mask_zero=True,
               weights=[embedding_matrix],
               trainable=False)(inputs)

x = Bidirectional(GRU(64, return_sequences=True))(embed)
x = GlobalMaxPool1D()(x)
x = Dense(16, activation="relu")(x)
x = Dropout(0.1)(x)
final_layer = Dense(1, activation="sigmoid")(x)

model=Model(inputs=inputs,outputs=final_layer)