GC-12 Ordinal Sentiment Analysis on Amazon Reviews
(Previous Titled - Text Based Rating Prediction)
Team: Bhawana Sharma & Samantha Russo
Course: IT Capstone Section 01 Spring 2020
Advisor : Dr. Ying Xie
Affiliation: Kennesaw State University
Embedded Files
Summary
Summary
- Predict accurate user rating based on an ordinal scale for Amazon product's review text.
- Utilize the Ordinal Hyperplane (OHPL) solution developed by Dr. Bob Vanderheyden and Dr. Ying Xie, along with Long Term Memory (LSTM) to determine if deep neural network will out perform other state-of-the-art alternatives. Deep learning has not been directly applied to the ordinal classification problem, until the creation of OHPL
- Existing work reduces ordinal classes to binary classes by grouping class labels and apply Machine Learning methods like Support Vector Machine (SVM) and Naive Bayes classifier to achieve only binary prediction like pain/no pain, positive/negative, true/false.
What is Ordinal Classification?
What is Ordinal Classification?
Ordinal classification is a type of regression analysis used for predicting an ordinal variable. Class variable value exists on an arbitrary scale where only the relative ordering between different values is significant. As an example, level for pain can range between extremely low to extremely high or facial age estimation ranging from baby to old. Ordinal Classification is most commonly used in applications including rating scales, medical classification, meaningful grouping of continues data, facial age estimation etc.
What is Sentiment Analysis?
What is Sentiment Analysis?
Surveys have found 40% of consumers form an opinion about a product by reading just one to three reviews. If a product has 1 to 10 reviews, the easiest thing for a consumer to do is read each review. However, what if the product has 10,000+ reviews! This is where sentiment analysis can provide deeper insight into a product.
- Improve product features
- Increase conversion rate
- Improve customer service
- Improve product communication and other marketing strategies
Page updated
Google Sites
Report abuse