Nowadays, user's review or rating is very important for the company. Due to the narrative text, it is difficult to design a feature engine to extract important features from those reviews. Deep learning method can extract and learn features automatically, which gives me the motivation to design an innovative deep learning model to identify user reviews.

I desinged an innovative CLSTM-Attention model which contains word embedding layer, convolution layer, LSTM layer, attention layer and softmax layer.

Input: sentences/reviews Output: Positive/negative

What it does

The model is used to classify the reviews like user reviews into positive and negative reviews.

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for User Review Identification

Share this project: