Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/travelxml/tensorflow-and-keras-sentiment-analysis-using-rnn

NLP: Sentiment Analysis using TensorFlow and Keras: This project implements a Recurrent Neural Network (RNN) to classify Amazon product reviews into different sentiment categories. It leverages text preprocessing, tokenization, and model training to predict the sentiment with high accuracy.
https://github.com/travelxml/tensorflow-and-keras-sentiment-analysis-using-rnn

deep-learning deep-neural-networks deeplearning keras keras-tensorflow nlp rnn rnn-tensorflow sentiment-analysis tenserflow

Last synced: 14 days ago
JSON representation

NLP: Sentiment Analysis using TensorFlow and Keras: This project implements a Recurrent Neural Network (RNN) to classify Amazon product reviews into different sentiment categories. It leverages text preprocessing, tokenization, and model training to predict the sentiment with high accuracy.

Awesome Lists containing this project

README 

        
# TensorFlow and Keras: Sentiment Analysis Using RNN



This project uses advanced machine learning techniques to analyze the sentiment behind customer reviews. Specifically, it focuses on Amazon product reviews and determines whether a review is positive, negative, or somewhere in between. The goal is to predict the star rating of a review just by analyzing the text.



### What Is Sentiment Analysis?



Sentiment Analysis is a way of using computers to automatically figure out if a piece of text (like a product review) expresses a positive, negative, or neutral sentiment. For example, if someone writes "I love this product!", the sentiment is positive. If they write "This product is terrible," the sentiment is negative.



### What Is RNN?



Recurrent Neural Networks (RNN) are a type of artificial intelligence (AI) designed to understand sequences, like sentences or time-series data. RNNs are especially good at processing language because they can remember what has been said earlier in a sentence while interpreting the later parts.



### What Did We Do?



1. **Collected Data**: We used a dataset of Amazon product reviews, which includes both the text of the review and the star rating given by the customer.



2. **Processed the Text**: Before feeding the text to our model, we needed to clean it up and convert it into a form that the computer can understand. This involved:

   - Breaking down the reviews into individual words (tokenization).

   - Converting these words into numbers (since computers work with numbers).

   - Making sure all reviews are the same length by padding them (adding zeros where necessary).



3. **Built the Model**: We created an RNN model using TensorFlow and Keras. This model was designed to:

   - Take in the processed text.

   - Learn patterns that indicate whether a review is likely to have a high or low rating.

   - Predict a rating (1 to 5 stars) for new reviews.



4. **Trained the Model**: We fed the model thousands of reviews along with their ratings. The model used this data to learn how to predict ratings on its own.



5. **Tested the Model**: After training, we tested the model on new reviews that it hadn't seen before to see how well it could predict the correct rating.



6. **Made Predictions**: We created a simple function that allows you to input a new review, and the model will predict the star rating for that review.



## Getting Started



### Prerequisites



To run this project, you will need:

- A computer with Python installed.

- Basic knowledge of how to run Python scripts.

- TensorFlow and Keras libraries, which are used for building and running the AI model.



### Installation



1. **Clone the Repository**: Download the project files from GitHub.



   ```bash

   git clone https://github.com/TravelXML/TENSORFLOW-AND-KERAS-SENTIMENT-ANALYSIS-USING-RNN.git

   cd TENSORFLOW-AND-KERAS-SENTIMENT-ANALYSIS-USING-RNN

   ```



2. **Install Dependencies**: Install all the necessary Python libraries.



   ```bash

   pip install -r requirements.txt

   ```



### Running the Project



1. **Open the Jupyter Notebook**: This project is built using Jupyter Notebook, an easy-to-use interface for running Python code.



   ```bash

   jupyter notebook rnn_az_senti_analysis.ipynb

   ```



2. **Follow the Steps**: The notebook guides you through each step, from processing the data to training the model and making predictions.



3. **Make Predictions**: You can test the model with your own reviews by running the prediction function provided.



   ```python

   rating = predict_review_rating("This product is fantastic!")

   print(f"The predicted rating is: {rating}")

   ```



### Example Output



- **Input**: "Worst product ever!"

- **Predicted Rating**: 1 star



- **Input**: "Amazing quality, highly recommend!"

- **Predicted Rating**: 5 stars



## Understanding the Results



After running the model, you’ll get predictions for the sentiment of reviews. The model will predict how many stars a review is likely to get based on the words it contains. This can be incredibly useful for businesses wanting to automatically sort or respond to customer feedback.



---



### A Few Simple Concepts:

- **Machine Learning**: Teaching a computer to recognize patterns in data.

- **Neural Network**: A computer system modeled after the human brain that learns from data.

- **Training**: The process of feeding data to a model so it can learn from it.

- **Prediction**: Using a trained model to guess outcomes on new, unseen data.



Happy Coding!