Scribd
Upload
129 views

Rating System Based On Text Review Using Sentiment Analysis: Final Presentation

The document describes a rating system that automatically generates star ratings for user reviews on an e-commerce website based on sentiment analysis. It proposes using recurrent neural networks like LSTM and GRU to analyze product reviews and detect sentiment to assign an appropriate rating. The system was implemented on a website called ShopIN that allows users to leave text reviews which are then analyzed by an LSTM model trained on Amazon product reviews to predict a corresponding star rating. Initial results showed the system was able to successfully analyze reviews and generate ratings in real-time to improve the customer experience.

Uploaded by

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

Rating System Based On Text Review Using Sentiment Analysis: Final Presentation

The document describes a rating system that automatically generates star ratings for user reviews on an e-commerce website based on sentiment analysis. It proposes using recurrent neural networks like LSTM and GRU to analyze product reviews and detect sentiment to assign an appropriate rating. The system was implemented on a website called ShopIN that allows users to leave text reviews which are then analyzed by an LSTM model trained on Amazon product reviews to predict a corresponding star rating. Initial results showed the system was able to successfully analyze reviews and generate ratings in real-time to improve the customer experience.

Uploaded by

Tintu Trm
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 20

RATING SYSTEM BASED ON

FINAL PRESENTATION
TEXT REVIEW USING
SENTIMENT ANALYSIS
PRESENTED BY,
UNDER THE GUIDANCE OF, 02. ABDUL RASHEED A
PROF.  PRASANTH R S 04. ABHINAND MANOHAR OP
ASSISTANT PROFESSOR, 46. ATHUL KRISHNA PS
DEPARTMENT OF INFORMATION TECHNOLOGY 48. SACHIN KRISHNA
     PROBLEM STATEMENT​

• A customer review is a review of a product or service made by


a customer who bought that product. Genuine reviews help to
better understand the customers and improve customer service. 
• Reviews give the customers detailed information about
the product. These are the statements that reflect
customer's perception or sentiment. 
• Sometimes in E-commerce systems we see people giving
contradicting text-reviews and star ratings or sometimes the user may
also forget to give the star rating.
• This will cause the product to have fake reviews.
 PROPOSED SOLUTION

• We develop a shopping website named ‘ShopIN’ which will have


the feature of automatically generating the rating for user’s text
reviews based on the positivity and negativity of the sentences.
• This system will detect sentiments from the customer feedback
reviews on a product and generate a rating accordingly
 PROPOSED
ARCHITECTURE
RECURRENT NEURAL NETWORK

 Recurrent Neural Networks (RNNs) is a popular algorithm used in


sequence models.

 Sequence models are the machine learning models that input or output
sequences of data.
ALGORITHMS /  In sentiment classification opinions expressed in a piece of text is
TOOLS USED categorized. Here the input is a sequence of words.

 RNN maintains internal memory, due to this they are very efficient for
machine learning problems that involve sequential data. 
    Mainly there are two types of RNNs.

 LSTM( Long Short-Term Memory )


 GRU ( Gated Recurrent Units )
LONG SHORT TERM MEMORY

 Both LSTM and GRU are used as a solution to short-term memory.


They have internal mechanisms called gates that can regulate the flow
of information.

 These gates can learn which data in a sequence is important to keep or


ALGORITHMS / throw away. By doing that, it can pass relevant information down the
long chain of sequences to make predictions.
TOOLS USED
LSTM & GRU

 GRU uses less training parameter and therefore uses less memory and
executes faster than LSTM whereas LSTM is more accurate on a
larger dataset. One can choose LSTM if we are dealing with large
sequences and accuracy is concerned, GRU is used when we have less
memory consumption and want faster results.
FOR FRONT END:
 HTML
 CSS
 JavaScript
 Bootstrap
ALGORITHMS /
TOOLS USED FOR BACK END:

 MySQL Database
 Python
 Flask

DATASET : Amazon Product Review Dataset


(Collected from Kaggle)
PACKAGES:
• Numpy
• Pandas
• Tensorflow
ALGORITHMS / • Sklearn
TOOLS USED • Matplotlib
• NLTK
• Keras
SRS SRS
SDD SDD
WEB DEVELOPMENT STEPS

1) Developed separate Login for Admins and Users


IMPLEMENTATION
2) Designed Home page and Product pages

3) Developed text Review system


ML STEPS

1) Load the dataset ( Amazon Reviews Dataset)

2) Tokenize the sentences

3) Word Embedding

IMPLEMENTATION 4) Create LSTM model


( Use appropriate activation functions )

5) Train the model

6) Export the model


1) Login page

There are separate login pages for Admins and Users.

Admin login credentials are prestored in the database.

RESULTS
2) Home page

There are separate home pages for Admins and Users.

RESULTS • Admin home page

Admin have option to add new products


to the website.
2) Home page

There are separate home pages for Admins and Users.

RESULTS
• User home page

User can open each product and


give review.
2) Add product page

For admins, they can add new product details.

RESULTS

Also, there is an option for admins to delete the product, they can see the reviews,
But can’t edit it.
3) Rating prediction page 

For admins, they can add new product details.

RESULTS Here users can type reviews and


corresponding star rating will be
generated on the right side of the
screen by calling the ML model.
In most of the E-commerce websites we see people giving
contradicting text-reviews and star ratings or sometimes the user may
also forget to give the star rating, to solve these kinds of contradictions
and issues a automatically generating the rating for user’s text reviews
based on the positivity and negativity of the sentences. This will
CONCLUSION significantly improve customer experience by saving their time, it is
possible to analyze large amount of text data and extract opinions from
them, that will help both the businesses and the customers to achieve
their product goals. 
[1] X. Zhang and Q. Yu, “Hotel reviews sentiment analysis based on
word vector clustering,” 2017 2nd IEEE International Conference on
Computational Intelligence and Applications (ICCIA). 
[2] Elli, Maria Soledad, and Yi-Fan Wang. "Amazon Reviews,
business analytics with sentiment analysis." 2016 
[3] Aashutosh Bhatt, Ankit Patel, Harsh Chheda, Kiran Gawande,
“Amazon Review Classification and Sentiment Analysis “, IJCSIT,
REFERENCES Vol. 6 (6), 2015, 5107- 5110. 
[4] Shaikh, Tahura, and Deepa Deshpande. "Feature Selection Methods
in Sentiment Analysis and Sentiment Classification of Amazon Product
Reviews.",(2016) 
[5] B. Samal, A. Behera, and M. Panda, “Performance analysis of
supervised machine learning techniques for sentiment analysis,” 2017
Third International Conference on Sensing, Signal Processing and
Security (ICSSS). 
THANK YOU

You might also like