Scribd
Upload
3 views8 pages

NLP Implementation

This document presents a study on sentiment analysis of customer reviews from the Flipkart dataset using the Naïve Bayes classifier algorithm. It details the preprocessing steps, methodologies including Count Vectorizer and TF-IDF for feature extraction, and discusses the application of different types of Naïve Bayes classifiers. The study highlights the importance of sentiment analysis in the e-commerce industry and aims to evaluate the accuracy of the models used.

Uploaded by

Talha Taha1786
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views8 pages

NLP Implementation

This document presents a study on sentiment analysis of customer reviews from the Flipkart dataset using the Naïve Bayes classifier algorithm. It details the preprocessing steps, methodologies including Count Vectorizer and TF-IDF for feature extraction, and discusses the application of different types of Naïve Bayes classifiers. The study highlights the importance of sentiment analysis in the e-commerce industry and aims to evaluate the accuracy of the models used.

Uploaded by

Talha Taha1786
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Sentiment Analysis on Flipkart Dataset

using Naïve Bayes Classifier Algorithm

TASAWAR ABBAS KHAN


[FA22-RCS-016]
TALHA WAJID
[FA22-RCS-024]
Introduction
• Sentiment analysis is a common technique for extracting subjective information
from textual data. .
• The textual opinions, attitudes, and feelings are extracted and evaluated using
computational algorithms.
• It is particularly useful in the e-commerce industry, where customer reviews
serve as an essential data source.
• This paper focuses on sentiment analysis of customer reviews in the Flipkart
dataset using Naive Bayes theorem and Count Vectorizer, TF-IDF, two
commonly used techniques.
Data Set
•Flipkart Product & Customer Reviews Dataset (Import from Kaggle Dataset)
•2,05,052 Instances, 6 Attributes
Preprocessing
• To clean the data, special characters, punctuation, and stop words removed.
• Remove Neutral Values because of less contribution.
Methodology
•Count Vectorizer is a technique for converting text data into a numerical form
that machine learning algorithms can process in natural language processing.
•TF-IDF Term (used for feature extraction )
•Stemming is the process of reducing a word to its base or root form.
•For Example, word "running" has been reduced to just "run," which is how it
was originally written.
Methodology (conti..)
Naïve Bayes Classifiers :Naive Bayes is a well-known sentiment analysis machine learning
algorithm. This probabilistic classifier uses the Bayes theorem to classify data.
◦ Naive Bayes, a probabilistic classifier,
◦ works by determining, based on the presence or absence of particular characteristics.
◦ whether a text belongs to a particular sentiment class—positive, negative.
Bernoulli Naïve Bayes: It's commonly used for binary classification tasks where the
features are binary variables, typically representing presence or absence of certain
features.
•Gaussian Naïve Bayes: it is suitable for continuous data and normally distributed within
each class.
Compare Models Accuracy
Thank YOU

You might also like