Skip to content

Latest commit

 

History

History
53 lines (38 loc) · 1.15 KB

README.md

File metadata and controls

53 lines (38 loc) · 1.15 KB

Sentiment Analysis

This project analyses the performance of different methods in the challenging task of sentiment analysis. In order to do so, Amazon Kindle data was obtained from (https://nijianmo.github.io/amazon/index.html).

Implementation

The implementation process consisted in four main steps:

  • Pre-processing: Sampling (to overcome the class imbalance issue) and text cleaning.
  • Exploratory Data Analysis (EDA): N-grams, word count, etc.
  • Classification: Used both classical machine learning methods (such as Gradient Boosting, SVMs and Gaussian Naive Bayes) and deep learning methods (e.g. BERT, RoBERTa, LSTMs)
  • Text generation

Enviroment

  • Python 3.6+
  • Pandas
  • Spacy
  • Tqdm
  • Spacymoji
  • Numpy
  • Sklearn
  • Imblearn
  • Matplotlib
  • Seaborn
  • Gensim
  • PyLDAvis
  • Logging
  • Nltk
  • Wordcloud
  • Torchbearer
  • PyTorch
  • Transformers

Install all python modules with

pip install -r requirements.txt

or if you have different versions of Python installed:

pip3 install -r requirements.txt

Authors