The Internet Movie Database (IMDB) movie reviews data set is a well-known collection of plain-text film reviews, labeled as positive or negative (polarity of the opinion). Convolutional Neural Networks (CNNs) are a type of neural networks where filters are applied to groups of adjacent data elements to detect local paterns, which can be useful in the field of text analysis to detect associations of words in sentences.
The aim of this work is to train a deep learning model with CNNs to classify the polarity of the reviews with high accuracy (as close as possible to the state of the art), analyzing and evaluating the architecture, trying to refine it with different experiments.
| Programming language | Python 3 |
| Author | Albert Espín |
| Date | March 2019 |
| Code license | MIT |
| Report license | Creative Commons Attribution, Non-Commercial, Non-Derivative |