二、外文原文资料：

Design and Implementation of Mobile-Learning System

for Environment Education

Keywords: Interfaces,Abstract classes,methods

Abstract.

Amid the growing need for efficient and automated educational agents for mobile learning systems, learner demand for customized coursewares is increasing. However, many m-Learning systems that have been studied as of late have yet to support a mobile learning course in a seamless manner, thus failing to meet learner expectations. One of the problems of such m-Learning systems is their weakness in motivating learners, in particular, in the continuous feedback of the course. This paper proposes a mobile learning scheduling system that provides environment-related educational courses for learners. The proposed system monitors and evaluates the learning outcomes of learners on a continual basis, and calculates degrees of accomplishment in learning activities that will be applied to the agent's scheduling in order to provide learners with appropriate courses. Such courses enable learners to experience vigorous learning activities through repetitive programs, according to their ability.

1 Introduction

The fast development of Internet has recently enabled the on-line lecture through the e-learning system, which is now became popular topics in the area of computer education system industry. As this e-learning system is spread widely to the public, the users demand more diverse education service, and that results facilitating study on applied education service being very active [1-3]. Since the agent and broker for the domestic and foreign education software are organized to meet the demands of the average public more rather than customized service for individual learner, it is very difficult to accommodate the various needs for knowledge and evaluated level for each and every individual [4,5]. Although tools to help interaction between learners had been supported in many ways, in instructor's perspective, it is very hard to provide the right course schedule and combinations by analyzing each learner status after facing all registered learners. Hence, agent who can deliver feedback such as effective way of learning, course formation and course schedule to learners is needed in this e-learning system [6-8]. Multi-agent will be proposed in this paper, who can provides the appropriate active course scheduling and feed back to learner after evaluating the learners education level and method.

2 Course Scheduling Multi-agent

2.1 Mobile-Learning System Structure

In this m-learning system, learner and CSMA (Course Scheduling Multi-Agent) are

connected via mobile interface, and through Mobile Interface (MI) the request and

transfer for course scheduling occurs between learners and CSMA. Learners study the

course provided by CSMA in this system.

All the information created by CSMA will be stored into the database and if required, it will be loaded by CSMA and used for reorganizing course. Learner's profile

and information obtained by their learning activity as well will be stored into database

via MI, then by CSMA it will be regenerated and stored again as necessary information to the learner such as learning achievement level, course, scheduling, evaluation

data, feedback and etc. Figure 1 shows the structure for proposed m-learning system.

The key component of CSMA consists of four agents, CRA (Course Recomposition Agent), LAA(Learning Accomplishment Agent), LEA (Learning Evaluation Agent) and FA (Feedback Agent). The CRA is delivered the information on the

degree of accomplishment of learning from the LAA and creates and provides a new

and most customized learner-oriented course. The LAA estimates the degree of learning accomplishment based on the test results from the LEA and tracks the effectiveness of learning. The LEA is carrying out learning evaluation at every stage. The FA

provides relevant feedback to learners in accordance with the learners profile and

calculated degree of accomplishment of learning.

3 Experiments and Evaluations

A total of 80 persons were selected among many persons to provide a same courseware in which 40 persons were finally selected to evaluate a CSMA- and PDA-based learning system. Table 1 is a summary of the experimental environments.

For convenience, let the 40 persons who learn according to general learning methods be referred to as Learner Group A, and the 40 persons who learn according to the CSMA learning method be referred to as Learner Group B. Both the general learning methods implemented according to different items and the elements of the CSMA learning method are identical. For the general learning methods, learners themselves can assign learning times, while the CSMA learning method provides learners with suitable learning times according to the CSMA's course scheduling algorithm. The subsection-specific weakness is demonstrated with the graph and figures and the final test degree so that learners can compare it with their target score. Learners under the target degree can begin the repetition program by the course schedule provided by CSMA. Figure 3 offers information on learning data and the degree of learning accomplishment in PDA format.

In the experiment, both Learner Groups A and B were allowed to learn under learning environments where, while the learning elements were the same, the learning methods were different. As a method for identifying the difference in evaluation results between the two groups, a chi-square test (a statistical methodology) was used. Formula (6) shows the chi-square distribution.

Probabilistic verifications are available through comparison between calculated values and the values of statistical distribution tables, as well as through identifying the difference between the two variances. Table 2 shows chi-square values.