GitHub / Dhrumil-Zion / Sentiments-Prediction-Using-NLP

Predicting customer sentiments from feedbacks for amazon. While exploring NLP and its fundamentals, I have executed many data preprocessing techniques. In this repository, I have implemented a bag of words using CountVectorizer class from sklearn. I have trained this vector using the LogisticRegression algorithm which gives approx 93% accuracy. I have found out the top 20 positive and negative feedback words from thousands how feedbacks. Also after processing this much I have automated the whole process with one function so that it can be used as generic for many machine learning algorithms. I have also tested another algorithm called DummyClassifier which gives an accuracy of around 84%. After that, I have executed the famous algorithm which is TF-IDF for NLP. I have combined TF-IDF with LogisticRegression which gives almost 93% accuracy but deep insights. Also, while working with data has solved the problem of imbalanced data through RandomOverSampler class from imblearn library.

JSON API: http://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Dhrumil-Zion%2FSentiments-Prediction-Using-NLP
PURL: pkg:github/Dhrumil-Zion/Sentiments-Prediction-Using-NLP

Topics: dummyclassfier, imbalanced-data, imblearn, logistic-regression, nlp-machine-learning, randomoversampler, sklearn, tfidf-text-analysis

• Files
• Readme
• Dependencies
• Releases