Title

Sentiment Analysis Using Machine Learning

Document Type

Oral Presentation

Campus where you would like to present

Ellensburg

Event Website

https://digitalcommons.cwu.edu/source

Start Date

16-5-2019

End Date

16-5-2019

Abstract

Sentiment analysis is the process of computationally evaluating spoken or written language to determine if the message that is being conveyed yields a positive, negative or neutral opinion. It is really important because nowadays, companies receive a massive amount of reviews about their products, brands, websites, customer service etc... and the way this information is handled can be critical for their success. They therefore, must find a way to process this information quickly and accurately so that they can respond to customers’ needs on time. Having this in mind, we developed a software using machine learning tools, that is able to predict if a text will have a positive or negative impact on the reader. We used the TF-IDF (term’s frequency-Inverse Document frequency) technique to pre-process the input text (convert it to numbers) and the MLP (multilayer-perceptron) to classify the reviews. We obtained a sample of 25000 movie reviews from the IMDB website, trained our model using 75% of them and tested it with the rest(25%). Following this process, we were able to correctly classify 91% of the reviews that were used for testing. The software is written in Python using the machine learning library ”Scikit-Learn.” It is ready to be used and is able to classify a large amount of reviews in a short period of time.

Winner, Outstanding Oral Presentation, School of Graduate Studies and Research.

Faculty Mentor(s)

Razvan Andonie

Department/Program

Computer Science

Presentation1.pdf (621 kB)
Slides for SOURCE 2019 presentation Nkouanga

Additional Files

Presentation1.pdf (621 kB)
Slides for SOURCE 2019 presentation Nkouanga

Share

COinS
 
May 16th, 12:00 AM May 16th, 12:00 AM

Sentiment Analysis Using Machine Learning

Ellensburg

Sentiment analysis is the process of computationally evaluating spoken or written language to determine if the message that is being conveyed yields a positive, negative or neutral opinion. It is really important because nowadays, companies receive a massive amount of reviews about their products, brands, websites, customer service etc... and the way this information is handled can be critical for their success. They therefore, must find a way to process this information quickly and accurately so that they can respond to customers’ needs on time. Having this in mind, we developed a software using machine learning tools, that is able to predict if a text will have a positive or negative impact on the reader. We used the TF-IDF (term’s frequency-Inverse Document frequency) technique to pre-process the input text (convert it to numbers) and the MLP (multilayer-perceptron) to classify the reviews. We obtained a sample of 25000 movie reviews from the IMDB website, trained our model using 75% of them and tested it with the rest(25%). Following this process, we were able to correctly classify 91% of the reviews that were used for testing. The software is written in Python using the machine learning library ”Scikit-Learn.” It is ready to be used and is able to classify a large amount of reviews in a short period of time.

Winner, Outstanding Oral Presentation, School of Graduate Studies and Research.

https://digitalcommons.cwu.edu/source/2019/Oralpres/174