AMAZON PRODUCT REVIEW RATING PREDICTION WITH CONVOLUTIONAL NEURAL NETWORK (CNN)
Sentiment analysis is the application of Natural Language Processing (NLP) to determine the sentiments of texts expressed by humans. One application of sentiment analysis that is popular in the e-commerce sector is sentiment classification. Sentiment classification is used to categorize customer rev...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/74370 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| id |
id-itb.:74370 |
|---|---|
| spelling |
id-itb.:743702023-07-12T07:53:23ZAMAZON PRODUCT REVIEW RATING PREDICTION WITH CONVOLUTIONAL NEURAL NETWORK (CNN) Zidane Faturrahman, Andika Indonesia Final Project Sentiment analysis, sentiment classification, deep learning, CNN, NLP INSTITUT TEKNOLOGI BANDUNG https://digilib.itb.ac.id/gdl/view/74370 Sentiment analysis is the application of Natural Language Processing (NLP) to determine the sentiments of texts expressed by humans. One application of sentiment analysis that is popular in the e-commerce sector is sentiment classification. Sentiment classification is used to categorize customer reviews of a product into positive reviews or negative reviews. The rating of the review is represented by a likert scale from 1 to 5. The Convolutional Neural Network (CNN) model is a deep learning model that can be used to clasify sentiments based on the effect of multiple pairwise words. This study aims to analyze the application of the CNN model in predicting ratings from product reviews on Amazon. Result showed that the CNN model could predict review’s ratings by 0.33 out of 1.0 F1 score and parameters that give the huge effect on CNN model are the number of filters, activation function, and also optimization algorithm for optimizing CNN parameters. text |
| institution |
Institut Teknologi Bandung |
| building |
Institut Teknologi Bandung Library |
| continent |
Asia |
| country |
Indonesia Indonesia |
| content_provider |
Institut Teknologi Bandung |
| collection |
Digital ITB |
| language |
Indonesia |
| description |
Sentiment analysis is the application of Natural Language Processing (NLP) to determine the sentiments of texts expressed by humans. One application of sentiment analysis that is popular in the e-commerce sector is sentiment classification. Sentiment classification is used to categorize customer reviews of a product into positive reviews or negative reviews. The rating of the review is represented by a likert scale from 1 to 5. The Convolutional Neural Network (CNN) model is a deep learning model that can be used to clasify sentiments based on the effect of multiple pairwise words. This study aims to analyze the application of the CNN model in predicting ratings from product reviews on Amazon. Result showed that the CNN model could predict review’s ratings by 0.33 out of 1.0 F1 score and parameters that give the huge effect on CNN model are the number of filters, activation function, and also optimization algorithm for optimizing CNN parameters. |
| format |
Final Project |
| author |
Zidane Faturrahman, Andika |
| spellingShingle |
Zidane Faturrahman, Andika AMAZON PRODUCT REVIEW RATING PREDICTION WITH CONVOLUTIONAL NEURAL NETWORK (CNN) |
| author_facet |
Zidane Faturrahman, Andika |
| author_sort |
Zidane Faturrahman, Andika |
| title |
AMAZON PRODUCT REVIEW RATING PREDICTION WITH CONVOLUTIONAL NEURAL NETWORK (CNN) |
| title_short |
AMAZON PRODUCT REVIEW RATING PREDICTION WITH CONVOLUTIONAL NEURAL NETWORK (CNN) |
| title_full |
AMAZON PRODUCT REVIEW RATING PREDICTION WITH CONVOLUTIONAL NEURAL NETWORK (CNN) |
| title_fullStr |
AMAZON PRODUCT REVIEW RATING PREDICTION WITH CONVOLUTIONAL NEURAL NETWORK (CNN) |
| title_full_unstemmed |
AMAZON PRODUCT REVIEW RATING PREDICTION WITH CONVOLUTIONAL NEURAL NETWORK (CNN) |
| title_sort |
amazon product review rating prediction with convolutional neural network (cnn) |
| url |
https://digilib.itb.ac.id/gdl/view/74370 |
| _version_ |
1822007378252398592 |