Sequence to Sequence Model Performance for Education Chatbot


  • Kulothunkan Palasundram Universiti Putra Malaysia
  • Nurfadhlina Mohd Sharef Universiti Putra Malaysia
  • Nurul Amelina Nasharuddin Universiti Putra Malaysia
  • Khairul Azhar Kasmiran Universiti Putra Malaysia
  • Azreen Azman Universiti Putra Malaysia



Education chatbot, natural language conversation, natural answer generation, question answering, sequence to sequence learning, Seq2Seq


Chatbot for education has great potential to complement human educators and education administrators. For example, it can be around the clock tutor to answer and clarify any questions from students who may have missed class. A chatbot can be implemented either by ruled based or artificial intel-ligence based. However, unlike the ruled-based chatbots, artificial intelli-gence based chatbots can learn and become smarter overtime and is more scalable and has become the popular choice for chatbot researchers recently. Recurrent Neural Network based Sequence-to-sequence (Seq2Seq) model is one of the most commonly researched model to implement artificial intelli-gence chatbot and has shown great progress since its introduction in 2014. However, it is still in infancy and has not been applied widely in educational chatbot development. Introduced originally for neural machine translation, the Seq2Seq model has been adapted for conversation modelling including question-answering chatbots. However, in-depth research and analysis of op-timal settings of the various components of Seq2Seq model for natural an-swer generation problem is very limited. Additionally, there has been no ex-periments and analysis conducted to understand how Seq2Seq model handles variations is questions posed to it to generate correct answers. Our experi-ments add to the empirical evaluations on Seq2Seq literature and provides insights to these questions. Additionally, we provide insights on how a cu-rated dataset can be developed and questions designed to train and test the performance of a Seq2Seq based question-answer model.

Author Biographies

Kulothunkan Palasundram, Universiti Putra Malaysia

KULOTHUNKAN PALASUNDRAM graduated from Universiti Kebangsaan Malaysia and received B.S in Computer Science (Hons) and Masters in I.T degrees in 1995 and 1998 respectively. He is currently pursuing the Ph.D. degree in Intelligent Computing in Universiti Putra Malaysia. His research interests include artificial intelligence, deep learning, big data, natural language processing and dialogue generation

Nurfadhlina Mohd Sharef, Universiti Putra Malaysia

Nurfadhlina Mohd Sharef is an Associate Professor at the Department of Computer Science and is currently the Deputy Director of Innovation in Teaching and Learning at the Centre for Academic Development (CADe) in UPM. She was the Head of Intelligent Computing Research Group at the Faculty of Computer Science and Information Technology, University of Putra (UPM), Malaysia. Dr. Fadhlina’s main research interest is in data science especially in solving sentiment analysis, question answering, chatbot and recommendation system problems. She has various experience in both academic and industrial projects involving development of intelligent computing methods especially adaptive and deep learning models for data science. Among her recent projects are the (i) deep learning based tensor factorization for recommender system , (ii) multi-objective particle swarm optimisation for breast cancer recurrence prediction, (iii) improvement of consistency and meaningfulness of a chatbot model, and (iv) multi-channel based transfer learning model for multi-class classification of tweets. She was also engaged in several consultation projects such as in the (a) online logistics aggregation web-based and mobile-based service, (b) pre-university intake requirements analysis and (c) the fuzzy aggregation based data analytics for security threat profiling from heterogeneous resources. In her portfolio at CADe her mantra is for InnoCreative Educators to innovate both conventional and technology as the important tools to deliver meaningful teaching and to accomplish Putra Learning Experience for the students. She is usually assigned to teach courses related to Artificial Intelligence and general computer science skills such as Data Mining, Intelligent Computing, and Programming I and II. In her teaching, she usually emphasizes on experiential learning and believes blended learning is the best method to learn and teach. She uses multiple teaching modalities to ensure engaging delivery and so that higher order thinking skills could be obtained. She has also conducted several training sessions locally and internationally relating to innovation in teaching and learning generally and e-learning specifically.

Nurul Amelina Nasharuddin, Universiti Putra Malaysia

NURUL AMELINA NASHARUDDIN received her Ph.D. from the Universiti Putra Malaysia, Malaysia in 2017. She is a senior lecturer at the Department of Multimedia, Faculty of Computer Science and Information Technology. He interests include Natural Languge Processing, Cross-language Information Retrieval

Khairul Azhar Kasmiran, Universiti Putra Malaysia

KHAIRUL AZHAR KASMIRAN received his Ph.D. from the University of Sydney, Australia in 2012. He is a senior lecturer at the Department of Computer Science, Faculty of Computer Science and Information Technology, Universiti Putra Malaysia, Malaysia. His interests include deep learning, reinforcement learning, per-formance engineering, formal verification and software development multi-task deep learning for multi-class tweets classification and deep tensor factorization model for recommendation system.

Azreen Azman, Universiti Putra Malaysia

AZREEN AZMAN is an Associate Professor at the Universiti Putra Malay-sia. He received a Diploma in Software Engineering from the Institute of Telecom-munication and Information Technology in 1997. Immediately, he was accepted directly to second year in Multimedia University, Malaysia to study Bachelor of In-forma tion Technology majoring in Information Systems Engineering. He completed his bachelor degree in 1999. After serving in the industry for a few years, he enrolled for a PhD in January 2003, studying Computing Science specializing in Information Retrieval in the University of Glasgow, Scotland and completed his study in Septem-ber 2007. His current research interests include information retrieval, text mining, natural language processing and intelligent systems. He serves as a committee mem-ber for the Malaysian Society of Information Retrieval and Knowledge Management (PECAMP) and the Malaysian Information Technology Society (MITS).




How to Cite

Palasundram, K., Mohd Sharef, N., Nasharuddin, N. A., Kasmiran, K. A., & Azman, A. (2019). Sequence to Sequence Model Performance for Education Chatbot. International Journal of Emerging Technologies in Learning (iJET), 14(24), pp. 56–68.