Sentiment analysis for airline services on Twitter using deep learning with word embedding / Mawada Mohamed Nour El Daim El Khalifa

Mawada Mohamed , Nour El Daim El Khalifa (2020) Sentiment analysis for airline services on Twitter using deep learning with word embedding / Mawada Mohamed Nour El Daim El Khalifa. Masters thesis, Universiti Malaya.

	PDF (The Candidate's Agreement) Restricted to Repository staff only Download (217Kb)
	PDF (Thesis M.A) Download (1364Kb)

Abstract

The use of social media platform in the airline industries have increased rapidly to allow analysis introduce the quality and performance of the services. The role of Sentiment Analysis (SA) is to classify people's opinions into different categories, such as positive and negative from text, using existing algorithms. However, existing approaches such as the Bag of Words (BOW) model is frequently used for text classification, where a document is mapped to a feature vector before the construction of the actual model, using machine learning techniques, like Logistical Regression and Support Vector algorithms. This problem has led to low accuracy in predicting Airline Services using Twitter data. Meanwhile, in recent years, Deep Learning algorithms for Sentiment Analysis has emerged as one of the most popular algorithms, which provides automatic feature extraction, rich representation capabilities, and better performance than most of the traditional learning algorithms. This research proposes deep learning with word embedding prediction model in order to improve the accuracy of the sentiment analysis of the airline services. The objectives of this research are as follows: to achieve the research aim, wherein firstly a taxonomy to classify different airline services models in the current literature based on sentiment classifiers is designed. Secondly, this research achieved the development of deep learning with a word embedding (DLWE) prediction model that extract features automatically through the layers in the model to classify and improve the prediction accuracy of sentiment analysis of the airline services. Thirdly, experiment to evaluate and compare the performance of the deep learning based sentiment classifier with some of the machine learning baselines has been conducted. For the experiment, the research used a public dataset that were extracted from the microblogging and airline services domain, while the training and testing process was performed using TensorFlow's library, for the evaluation process, classification accuracy and confusion matrix were employed. Regarding the results, the highest prediction accuracy achieved is 82.61% by using the word embedding model as a features extractor, and deep neural network learning as a classifier, as to when compared to the existing benchmark which used SVM and LG machine learning classifiers and word to vector model as a features extractor with 72% prediction accuracy. Thus, the result obtained from the proposed deep learning with word embedding model, can be used to help improve the sentiment analysis accuracy of airline services and other related fields of future research. Keywords:

Actions (For repository staff only : Login required)