Juei-Nan Chen

Applying dynamic fuzzy Petri Net to web learning system

This investigation presents a DFPN (Dynamic Fuzzy Petri Net) model to increase the flexibility of the tutoring agents behaviour and thus provide a learning content structure for a lecture course. The tutoring agent is a software assistant for a single user, who may be an expert in an e-Learning course. Based on each learners behaviour, the tutoring agent derives a different learning content structure, and then maps it onto a SCORM activity tree structure. SCORM is the most widely accepted standard e-Learning application. However, it is too complex to enable the lecturers to apply SCORM Sequencing and Navigation in developing a course. Accordingly, this investigation applies DFPN to provide a graphical editing interface, to enable the tutoring agent to help each learner reach his (or her) learning target.

An intelligent human-expert forum system based on fuzzy information retrieval technique

The forum system is useful for sharing knowledge and help-seeking. However, existing forums often have the problem that some questions remain unanswered. This study proposes an intelligent human-expert forum system to perform more efficient knowledge sharing. The system uses fuzzy information retrieval techniques to discover important discussion knowledge and actively invites human-experts who might be capable of answering the question to participate in the discussion. The selection of human-experts employs the Expert-Terms Correlation Matrix, which stores the knowledge strength of human-experts. Moreover, the forgetting curve is adopted into our forum system for modeling the variations in memory strength. The experiment uses three different categories of discussions to be the testing data, and the performance study shows acceptable results on discussion searching and expert discovery.

NP-miner: a real-time recommendation algorithm by using web usage mining

Web usage mining is widely applied in various areas, and dynamic recommendation is one web usage mining application. However, most of the current recommendation mechanisms need to generate all association rules before recommendations. This takes lots of time in offline computation, and cannot provide real-time recommendations for online users. This study proposes a Navigational Pattern Tree structure for storing the web accessing information. Besides, the Navigational Pattern Tree supports incremental growth for immediately modeling web usage behavior. To provide real-time recommendations efficiently, we develop a Navigational Pattern mining (NP-miner) algorithm for discovering frequent sequential patterns on the proposed Navigational Pattern Tree. According to historical patterns, the NP-miner scans relevant sub-trees of the Navigational Pattern Tree repeatedly for generating candidate recommendations. The experiments study the performance of the NP-miner algorithm through synthetic datasets from real applications. The results show that the NP-miner algorithm can efficiently perform online dynamic recommendation in a stable manner.

Automatic leveling system for e-learning examination pool using entropy-based decision tree

In this paper, we propose an automatic leveling system for e-learning examination pool using the algorithm of the decision tree. The automatic leveling system is built to automatically level each question in the examination pool according its difficulty. Thus, an e-learning system can choose questions that are suitable for each learner according to individual background. Not all attributes are relevant to the classification, in other words, the decision tree tells the importance of each attribute.

ANTS: agent-based navigational training system

There is an explosive growth of e-learning trend during the last few years. More and more learning resources are generated by different purposes in the world. How to make the learning resources to be sharable and reusable is a key factor in e-learning environment. This paper presents a system framework of Agent-based Navigational Training System (ANTS) to facilitate lecturers and learners achieving their works. Besides, we apply the model of Dynamic Fuzzy Petri Net (DFPN) and the intelligent agent into the system to assist learners. The intelligent agent can dynamically generate the learning path for each learner. Moreover, the system is compatible to Sharable Content Object Reference Model (SCORM) which is the most acceptable e-Learning system developing standard. Accordingly, the learning resources could be sharable and reusable in any platforms which are compatible to SCORM standard.

An intelligent database system for composite material selection in structural design

Abstract This paper describes the development of a prototype intelligent database system for composite material selection in structural design. This intelligent system integrates the expert system with the database system to provide decision-making support systems that exhibit some forms of intelligence. The overall architecture of this system is illustrated. The present capabilities of this system are discussed and demonstrated with an example problem.

Cloud computing in Taiwan

Cloud computing is changing the computing environment: scalable, virtualized resources are increasingly provided as services over the Internet. Taiwan is also changing, transforming itself from a hardware manufacturing island into a cloud village offering both services and resources.

Context-sensitive content representation for mobile learning

Mobile learning means that the learning contents can be displayed anytime, anywhere, and with any kind of presenting device. Learning Content Management Systems (LCMSs) usually provide convenient authoring tools to help instructors to construct their learning contents, which may include static document such as powerpoint, word, pdf document and dynamic multimedia document such as video and audio files, and then integrate these learning contents to provide learners with proper contents rendering through access devices. However, most of LCMSs are based on desktop computer environments, rather than mobile devices. Context-Sensitivity is an application of software systems ability to sense and analyze context from various sources. In this paper, we develop a Context-Sensitive Middleware (CSM) for LCMS to transform the same learning contents to different mobile devices, so mobile learning can be supported.

Agent-Based Web Learning System Applying Dynamic Fuzzy Petri Net

This paper presents an Agent-Based Web Learning System to facilitate each learner achieving his learning target. Agent is a software assistant for single user, which can be an expert in specific domain. The agent can generate a learning sequence to the learner based on DFPN (Dynamic Fuzzy Petri Net). Besides, we suggest that each course should define the study intensity function to normalize different exercise grade criteria. In DFPN, each learner should receive different learning suggestions based on evaluation degree of truth of any proposition. Therefore, ABWLS can make students feel that ”everyone is different”, and achieve one-to-one learning effect.

Extended real-time learning behavior mining

Based on our previous work (Y. H. Kuo et al., 1999), learning patterns can be discovered and recommended to learners. This paper extends the proposed problem to handle the questionable mining results. According to the learning patterns discovered by using learning histories, it happened whenever the learners have ineffective learning behaviors, and we define them as questionable mining results. These ineffective behaviors may induce the bias suggestions. Therefore, we propose a candidate sequence set generation process to take care the stumble learning behavior.