Dissertations

    1. Mum - Postdoctoral dissertation in Mathematics and Statistics (2014 - 2016): The research dissertation composed as the book Mathematical Approaches to User Modeling focuses on innovative probabilistic and statistical works applied into user modeling, data mining, and machine learning.

    2. The dissertation is published in Journals Consortium.

    3. Hudup - Postdoctoral dissertation in computer science (2013): A framework of e-commercial recommendation algorithms. Abstract - The recommender framework dedicated to scientists and software developers who create or deploy recommendation solutions and algorithms in e-commerce and e-learning. Hudup is composed of three modules:

        • The infrastructure to setting up recommendation algorithms.

        • The evaluation system to measure recommendation algorithms according to metrics.

        • The simulation environment to execute and test recommendation solutions and algorithms before deploy them in real-time applications.

    4. Zebra - PhD dissertation in computer science (2009): A User Modeling System for Adaptive Learning. Abstract - Nowadays modern society requires every citizen always updates and improves her / his knowledge and skills necessary to working and researching. E-learning or distance learning gives everyone a chance to study at anytime and anywhere with full support of computer technology and network. Adaptive learning, a variant of e-learning, aims to satisfy the demand of personalization in learning. The adaptive learning system (ALS) is defined as the computer system that has ability to change its action to provide learning content and pedagogic environment/method for every student in accordance with her/his individual characteristics. Therefore, the ultimate goal of this research is to give the best support to learners in their learning path and this is an enthusiastic contribution to research community. Learners’ information and characteristics such as knowledge, goal, experience, interest, background, etc are the most important to adaptive system. These characteristics are organized in structure so-called learner model (or user model) and the system or computer software that builds up and manipulates learner model is called user modeling system (or learner modeling system). This research proposes a learner model that consists of three essential kinds of information about learners such as knowledge, learning style and learning history. Such three characteristics form a triangle and so this learner model is called Triangular Learner Model (TLM). The ideology of TLM is that user characteristics are various and only some information is really necessary to adaptive learning and an optimal user modeling system should choose essential information relating to user’s study to build up learner model.

    5. AGmagic - Master dissertation in computer science (2005): An image searching framework. Abstract - The research build up the image searching framework named AGmagic. The framework AGmagic uses the method Markov Model Mediator proposed by Professors Mei-Ling Shyu, Shu-Ching Chen, Min Chen, Chengcui Zhang, Kanoksri Sarinnapakorn to combine low-level features and high-level semantics contents of images in order to find images. Low-level features were extracted from segmentation, color histogram, color gradient, centroid. High-level semantic contents learned from user feedback are measures that reflect personal subjective feeling about the similarity among images. The combination of low-level features and high-level semantic contents gives out excellent result in searching images. Moreover AGmagic framework implements very successfully the segmentation algorithm proposed Prof. Doan, Van C.