Chatbot Based Customer Service Model Using Deep Learning the Case of Ethiopian Airlines
No Thumbnail Available
Date
2021-07-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Chatbot systems implemented for different purposes and plays significant role in terms of accessibility, reliability, and offers cost efficient auto services. The usage of chatbots grown rapidly in various fields in recent years, including Marketing, Supporting Systems, Education, Health Care, Cultural Heritage, and Entertainment. Therefore, we are motivated to design, develop, and implement automated Deep Learning based chatbots for Ethiopian airlines customer services. The reason to select Ethiopian airlines is even though it has a best customer service currently, the chatbot service will enhance improving its services more. This study aimed on designing and implementing a chatbot based model using deep learning methods which can facilitate customer service for enhancing Ethiopian airlines services. For this study, 30,000 question and answer pair statements has been collected from Ethiopian Airlines FAQ and from Kaggle websites. The collected documents have been passed through the appropriate data preparation. The dataset has split into 80% for training and 20% for testing sets. The researcher applied two different neural network techniques. The two neural network techniques experimented in this research are Long Short-Term Memory (LSTM) techniques and Convolutional Neural Network (CNN). To evaluate the performance of each technique, the researcher used various performance evaluation metrics such as Precession, Recall, F-score, Accuracy. The feature extraction techniques used for neural network techniques are word embedding, bag of words and word2vec methods. The evaluated Neural network techniques accomplished accuracy for LSTM 83.25% and CNN 85.20%. According to the performance result from the techniques applied, the CNN technique achieved better accuracy compared to LSTM and we applied CNN to deploy our model.
Description
Keywords
Chatbot, Deep Learning, Neural Network, Feature Extraction, Word2Vec