Sentiment Analysis - Reviews YELP

In order to encourage the learning of artificial intelligence, YELP provides business datasets, assessments and user data for use in personal, educational and academic purposes.

In this one, we will focus on evaluation data, where the goal is to develop a classifier that can distinguish "positive" and "negative" messages. For this task we will use several Deep Learning techniques, such as:

Perceptron
LSTM
BERT

Technology Used

Jupyter Notebook
Docker
Angular 10
Python

New Features!

Classify positive and negative using Perceptron
API for expose model
Create of the angular design architecture
Create of the docker architecture

[Pytorch] - Framework by Deep Learning
[Numpy] - awesome web-based text editor
[Jupyter Notebook] - Markdown parser done right. Fast and easy to extend.

Trainning Model

The project require Pytorch v1.2.2+ to run.

After install dependencie, run file train.py.

$ cd model
$ python train.py

Todos

Implement BERT transformer
Create Web PAge

License

MIT