Embedded Files
Sentiment Analysis and Topic Modelling
Sentiment Analysis and Topic Modelling
Used a kaggle Flipkart Product Reviews Dataset. This project consists of Exploratory Data Analysis for free text data from the reviews. Sentiment analysis was performed using DistilBERT model from huggingface by fine-tuning it for the product reviews data. Topic modelling was performed using BERTopic with a spacy model for embeddings.
Highlights
Highlights
Sentiments (Positive, Negative, Neutral) are analysed as per the price range of products.
Wordclouds analysed for product names and summary of review features.
DistilBERT model fine-tuned on the data. Recall: 0.94 Precision: 0.94
Topic Modelling performed for product names and summary of review features. Topics visualized in a 2D space for inter-topic distance
Gallery
Gallery
Sentiment Count
Word Cloud
Topic Modelling Product Names
Page updated
Google Sites
Report abuse