Hi, I'm Mayuri!
Techie, Data Enthusiast, Insight Analyst
- Dimensionality reduction for sentiment analysis using pre-processing techniques
- Conducted an in-depth analysis of various text pre-processing techniques for sentiment analysis, including Handling Expressive Lengthening, Emoticons Handling, HTML Tags Removal, Slangs Handling, Punctuations Handling, Stopwords Removal, Stemming, and Lemmatization, using Python and libraries like NLTK, scikit-learn, BeautifulSoup, pytypo, re (Regular Expressions), WordNet and bs4.
- Developed a robust sentiment analysis model using Python's Random Forest Classifier, evaluating the impact of different combinations of pre-processing techniques on accuracy with the "Bag of Words Meets Bags of Popcorn" dataset from Kaggle, employing 10-fold cross-validation for reliable performance assessment.
- Demonstrated that applying Python-based pre-processing techniques significantly enhances sentiment analysis accuracy compared to using unprocessed data, providing valuable insights for improving text analysis in business contexts.
