USER AND APPLICATION INTERFACE

FOR A DISTRIBUTED E-LEARNING SYSTEM

Takao Kawamura, Kazuo Kuramochi

Faculty of Engineering, Tottori University, 4-101 Koyama-Minami, Tottori, Japan

Junya Kishida, Shinichi Motomura

Graduate School of Engineering, Tottori University, 4-101 Koyama-Minami, Tottori, Japan

Kazunori Sugahara

Faculty of Engineering, Tottori University, 4-101 Koyama-Minami, Tottori, Japan

Keywords:

Distributed System, P2P, Mobile Agent, Web-based training, e-Learning.

Abstract:

In this paper, the user and application interface of our previously proposed e-Learning system is presented to

demonstrate how user interfaces for distributed systems should be composed. The proposed e-Learning system

has two distinguishing features. Firstly, it is based on P2P architecture for scalability and robustness. Secondly,

all exercises in the system are not only data but also agents so that they have functions, such as scoring

user’s answers, telling the correct answers, and showing some related information without human instruction.

The interface provides three kinds of communication: communication between a user and an exercise agent,

communication between an exercise agent and support applications, and communication between a user and

other users. The interface consists of two parts: an interface agent and a plugin for Firefox web browser. An

interface agent hides the existence of the P2P network and other agents from the Firefox plugin and users. The

developed e-Learning system including the interface is examined by expe

riments in classroom.

1 INTRODUCTION

Nowadays, e-Learning systems are very popular ev-

erywhere, such as college education, corporate educa-

tion, and community education. The term e-Learning

covers a wide set of applications and processes, such

as Web-based training (hereafter we abbreviate as

WBT), computer-based training, virtual classrooms,

and digital collaboration. We are concerned with

asynchronous WBT that allows the learner to com-

plete the WBT on his own time and schedule, without

live interaction with the instructor.

Although a large number of studies have been

made on asynchronous WBT (Nishita et al., 2002;

Homma and Aoki, 2003; Helic et al., 2005), all

of them are based on the client/server model. The

client/server systems generally lack scalability and

robustness. In the recent years, P2P research has

grown exponentially. Although the current P2P sys-

tems are famous for ﬁle sharing, and the consequent

legal problems, P2P systems are gradually proving

to be a very promising area of research because

they have potential for offering a decentralized, self-

sustained, scalable, fault tolerant and symmetric net-

work of computers providing an effective balancing

of storage and bandwidth resources.

We have proposed and implemented a distributed

e-Learning system based on P2P architecture (Kawa-

mura and Sugahara, 2005; Kawamura et al., 2005a;

Motomura et al., 2006c) using Maglog that is a

Prolog-based framework for building mobile multi-

agent systems we have also developed (Kawamura

et al., 2005b; Motomura et al., 2006a; Motomura

et al., 2006b). The proposed e-Learning system has

two distinguishing features. Firstly, it is based on P2P

architecture and every user’s computer plays the role

of a client and a server. Namely, while a user uses the

proposed e-Learning system, his/her computer (here-

after we refer to such a computer as a node) is a part

of the system. It receives some number of contents

from another node when it joins the system and has

responsibility to send appropriate contents to request-

ing nodes. Secondly, each exercise in the system is

not only data but also an agent so that it has functions,

357

Kawamura T., Kuramochi K., Kishida J., Motomura S. and Sugahara K. (2007).

USER AND APPLICATION INTERFACE FOR A DISTRIBUTED E-LEARNING SYSTEM.

In Proceedings of the Third International Conference on Web Information Systems and Technologies - Internet Technology, pages 357-363

DOI: 10.5220/0001273303570363

 SciTePress

such as scoring user’s answers, telling the correct an-

swers, and showing some related information without

human instruction.

In this paper, we present the user and application

interface (hereafter we abbreviate as UAI) of the pro-

posed e-Learning system to demonstrate how user in-

terfaces for distributed systems should be composed.

This paper is organized in 6 sections. The pro-

posed e-Learning system is described in Section 2.

We describe three functions of the UAI in Section 3

and an implementation of the UAI in Section 4, re-

spectively. In Section 5 experimental results are pre-

sented. Finally, some concluding remarks are drawn

in Section 6.

2 PROPOSED E-LEARNING

SYSTEM

2.1 Overview

As mentioned in the previous section, we focus on

asynchronous WBT, that is to say, a user can connect

to the proposed e-Learning system anytime and any-

where he/she wants. Once connection is established,

the user can obtain exercises one after another through

specifying categories of the required exercises. User’s

answers for each exercise are scored as correct or in-

correct right away. Related information may be pro-

vided for each answer, which can be viewed when the

correct answer is shown.

While a user uses the proposed e-Learning system,

his/her computer is a part of the system. Namely, it

receives some number of categories and exercises in

them from another node when it joins the system and

has responsibility to send appropriate exercises to re-

questing nodes.

The important point to note is that the categories

a node has are independent of the categories in which

the node’s user are interested as shown in Figure 1.

Figure 1 illustrates that user A’s request is forwarded

at ﬁrst to the neighbor node, next forwarded to the

node which has the requested category.

2.2 P2p Network

All exercises in the proposed system are classi-

ﬁed into categories, such as “English/Grammar”,

“Math/Statistics”, and “History/Rome”, etc.

When the proposed system bootstraps, one initial

node has all categories in the system. When another

node joins the system, it is received certain number

of categories from the initial node. The categories are

Figure 1: Proposed e-Learning system.

distributed among all nodes in the system according

as nodes join the system or leave the system.

We would like to emphasize that in existing P2P-

based ﬁle sharing systems, such as Napster (Napster,

2003), Gnutella (Gnutella, 2001), and Freenet (Clarke

et al., 1999) each shared ﬁle is owned by a particu-

lar node. Accordingly, ﬁles are originally distributed

among all nodes. On the other hand, the categories

in the proposed system are originally concentrated.

Consequently, when a new node joins the system, not

only location information of a category but the cate-

gory itself must be handed to the new node. Consider-

ing that, the P2P network of the proposed system can

be constructed as a CAN (Ratnasamy et al., 2001).

The CAN has a virtual coordinate space that is

used to store (key, value) pairs. To store a pair

, V

), key K

is deterministically mapped onto a

point P in the coordinate space using a uniform hash

function. The corresponding (key, value) pair is then

stored at the node that owns the zone within which

the point P lies. In the proposed system, we let each

category be a key and let a set of exercises belonging

to the category be the corresponding value.

Our P2P network is constructed with 2-

dimensional coordinate space [0,1] × [0,1] to

store exercise categories, as shown in Figure 2.

The ﬁgure shows the situation that node C has

just joined the system as the third node. Before

node C joins, node A and node B shared the whole

coordinate space half and half. At that moment, node

A managed “Math/Geometry” ,“Math/Statistics”,

and “History/Rome” categories and node B managed

“English/Grammar”, “English/Reader and “His-

tory/Japan” categories, respectively. When node C

WEBIST 2007 - International Conference on Web Information Systems and Technologies

358

joins the system, it is mapped on a certain coordinate

space according to a random number and takes on

corresponding categories from another node. For

example, in the case of Figure 2, node C takes on the

“History/Japan” category from node B and exercises

of the category move to node C.

Figure 2: P2P network.

2.3 Mobile Agents

Generally, in addition to service to show an exercise,

a WBT server provides services to score user’s an-

swers, to tell the correct answers, and show to some

related information about the exercise. Therefore, for

the proposed system that can be considered a dis-

tributed WBT system, it is not enough that only ex-

ercises are distributed among all nodes. Functions to

provide the above services also must be distributed

among all nodes. We adopt mobile agent technology

to achieve this goal. Namely, an exercise is not only

data but also an agent so that it has functions, such

as scoring user’s answers, telling the correct answers,

and showing some related information about the ex-

ercise.

In addition, mobile agent technology is applied to

realize the migration of categories, that is, each cate-

gory is also an agent in the proposed system.

3 USER AND APPLICATION

INTERFACE

There are three types of possible communications be-

tween for using the distributed e-Learning system. In

this section, three functions of the UAI to realize these

communications are described.

3.1 Function for Communication

Between a User and an Exercise

Agent

Although each exercise agent has a question and func-

tions, such as scoring user’s answers, telling the cor-

rect answers, and showing some related information

about the exercise, it cannot communicate with users

directly. All agents in the proposed system communi-

cate with other agents or applications through ‘ﬁeld’s

provided by Maglog framework. A ﬁeld is a kind of

preemptive queue. When agent-A intends to send a

message to agent-B, agent-A queues the message onto

the ﬁeld owned by agent-B. Since agent-B executes a

message dispatch loop like all other agents, it reads

the message written by agent-A sometime. The UAI

therefore has a function to convert user’s request into

messages and to write them onto the ﬁelds of exercise

agents, and vice versa.

3.2 Function for Communication

Between an Exercise Agent and

Applications

An exercise agent may need some support applica-

tions to score user’s answer. For example, when the

correct answer is “x

+x+y”, other forms of answers,

such as “y+ x+x

” or “x+ x

+ y” answered in a text

box should be scored as correct. However, it is impos-

sible for exercise agents to score all possible forms of

the correct answer as correct without a computer al-

gebra system. An exercise requiring C programming

is another example. An exercise agent must compile

and execute a user’s answer to compare the output of

it with the contents of the correct answer. Since ex-

ercise agents migrate between nodes, it is an infeasi-

ble idea that each exercise agent includes a computer

algebra system or a C compiler. Instead, in the pro-

posed system, exercise agents utilize support applica-

tions, such as a computer algebra system or a C com-

piler prepared on each node. The UAI therefore has a

function for communication between exercise agents

and support applications.

USER AND APPLICATION INTERFACE FOR A DISTRIBUTED E-LEARNING SYSTEM

359

3.3 Function for Communication

Between a User and Other Users

Users may need to be advised to understand correct

answers. Existing e-Learning systems often provide

BBS (Bulletin Board System) to deal such situation.

However, we don’t adopt BBS because it is not suit-

able for distributed systems. In addition, questions on

BBS are neglectable. Instead, we let the UAI have a

function for users to help each other by means of chat.

4 IMPLEMENTATION OF THE

USER AND APPLICATION

INTERFACE

We have implemented the UAI consisting of two

parts: an interface agent and a plugin for Firefox web

browser (Mozilla Corporation, 2005). Users try ex-

ercises by using Firefox with the plugin that does not

know the existence of the P2P network and any agents

except an interface agent. In this section, we describe

the architecture of a node at ﬁrst. Next we describe

the two parts of the UAI, respectively.

4.1 Node Architecture

As shown in Figure 3, the following agents exist on

each node. An agent server in this ﬁgure is a runtime

environment for agents.

Node Agent There is one node agent on each node.

It manages the zone information of a CAN and

forwards messages to the category agents in the

node.

Exercise Agent Each exercise agent has a question

and functions to score user’s answers, tell the cor-

rect answers, and show some related information

about the exercise. These data are formatted in

HTML.

Category Agent Each category agent stands for a

unit of a particular subject. It manages exercise

agents in itself and sends them to the requesting

node.

Interface Agent There is one interface agent on each

node. It is an agent part of the UAI.

User Agent There is one user agent for each user. A

user agent manages user information that include

computer, online/ofﬂine status, and study log.

Group Agent A group agent exists for grouping user

agents as a category agent exists for grouping ex-

ercise agents. Currently, users are grouped by the

ﬁrst letter of their login name, so that there are

26 group agents corresponding to one letter in the

English alphabet.

Figure 3: Agents in a node.

4.2 Firefox Plugin

Users try to answer exercises by using Firefox with

the plugin. Figure 4 is a screenshot of the main

user interface window generated by the plugin. Cate-

gories are classiﬁed into a hierarchical tree structure,

as shown in the category pane of Figure 4. By click-

ing the left button of a mouse on a category, Users

can select the category. An exercise belonging to the

category is obtained from some node and is shown in

the exercise pane by clicking the exercise button on

the menu bar. Users can require to score their answer

anytime by clicking the score button. Correct answers

and related information about the exercise are shown

by clicking the answer button. In addition to them,

users can request some advise about the exercise by

clicking the chat button.

4.3 Interface Agent

When a user tries an exercise, in addition to the agent

server, two types of processes run as shown in Figure

5. One type is for Firefox with the plugin for the UAI.

The other type is for support applications. They help

exercise agents to score students’ answers. There is

a wrapping Java class for each support application to

be utilized by an interface agent.

An interface agent enables three kinds of commu-

nication as follows:

1. Communication between a user and an exercise

agent.

WEBIST 2007 - International Conference on Web Information Systems and Technologies

360

Figure 4: Main user interface window of the proposed e-

Learning system.

Figure 5: User and application interface and other compo-

nents in a user studying node.

The Firefox plugin converts user’s request into

messages and writes them onto a ﬁled of an in-

terface agent via XML-RPC (Winer, 1998) using

jsolait (Kollhof, 2004).

After that, the plugin waits until the result of a

request is written on the same ﬁeld by the inter-

face agent. An interface agent forwards the re-

ceived messages to an appropriate agent: A re-

quest for obtaining an exercise is forwarded to the

node agent on the user’s computer, and requests

for being scored and for obtaining the correct an-

swer and related information are forwarded to the

exercise agent.

2. Communication between an exercise agent and

support applications.

All agents on Maglog, which are written in a

Prolog-like language, are translated into Java

source codes by our Maglog translator, and then

compiled into Java classes by the Java compiler

to being executed. An agent runs as a thread in

an agent server which is also implemented using

Java. Thereby agents can utilize any Java object,

i.e., an interface agent, which works as a proxy

of an exercise agent, can utilize a support appli-

cation for which a corresponding wrapping Java

class is prepared. A wrapping Java class utilizes

the standard Java class

Process

to invocate a sup-

port application and to communicate with it using

standard input/output.

3. Communication between a user and other users.

When a user wants to be advised about an exer-

cise, the following steps are performed.

(a) The interface agent asks the exercise agent the

potential advisers.

(b) The exercise agent tells a list of login names of

the users who gained a good score on itself to

the interface agent.

agent whether or not each user in the list is

logging-in the system. If a user is logging-in,

the IP address of the user’s computer is also

asked.

(d) The interface agent asks each user who is

logging-in the system and gained a good score

on the exercise in random order until one user

offers his/her service.

(e) The interface agent establishes a chat channel

between the two users.

5 EXPERIMENTS

The developed e-Learning system including the user

and application interface was examined by experi-

ments in classroom. For the experiments, 60 PCs with

Intel Pentium4 3.0GHz processor and 1GB of RAM

are connected via 1000Base-T network. All the PCs

were running on GNU/Linux (kernel version is 2.6.0)

operating system. The version of Java language run-

time environment was 1.4.2.

Figures 6 and 7 show, respectively, the result of

scoring, and the correct answer and related informa-

tion of the exercise.

USER AND APPLICATION INTERFACE FOR A DISTRIBUTED E-LEARNING SYSTEM

361

Figure 6: User’s answers are scored as correct or incorrect

by clicking the score button.

Figure 7: Correct answers and related information are

shown by clicking the answer button.

Figure 8 shows an example result of scoring us-

ing a computer algebra system, Maxima (Maxima

Project, 2000). In this example, the exercise agent

requests Maxima through an interface agent to com-

pare its correct answer with a user’s answer. Maxima

returns a subtraction of two expressions. The exer-

cise agent scores the user’s answer as correct or in-

correct whether the result of the subtraction equals

0 or not. Consequently, as shown in Figure 8, both

“cosx− sinx” and “− sinx+ cosx” are scored as cor-

rect.

As shown in Figure 9, while a chat session is in

progress, same windows are opened on the both two

users’ screen.

6 CONCLUSION

Since existing asynchronous WBT systems are based

on the client/server model, they have problems of

Figure 8: Example of scoring using a computer algebra sys-

tem.

Figure 9: Chat windows.

scalability and robustness. The proposed e-Learning

system solves these problems in decentralized manner

through both P2P technology and mobile agent tech-

nology.

In this paper, the user and application interface

of the proposed e-Learning system are described to

demonstrate how user interfaces for distributed sys-

tems should be composed. It realizes three kinds of

communication: communication between a user and

an exercise agent, communication between an exer-

cise agent and support applications, and communica-

tion between a user and other users. The user and

application interface consists of two parts: an inter-

face agent and a plugin for Firefox web browser. An

interface agent hides the existence of the P2P network

and other agents from the Firefox plugin and users.

The developed e-Learning system including the

user and application interface was examined by exper-

iments. More practical experiments are now planned

WEBIST 2007 - International Conference on Web Information Systems and Technologies

362

at our university to conﬁrm the effectiveness of the

proposed e-Learning system.

REFERENCES

Clarke, I., Sandberg, O., Wiley, B., and Hong,

T. W. (1999). Freenet: A distributed anony-

mous information storage and retrieval system.

http://freenetproject.org/freenet.pdf.

Gnutella (2001). Gnutella.com.

http://welcome.to/gnutella/.

Helic, D., Krottmaier, H., Maurer, H., and Scerbakov, N.

(2005). Enabling project-based learning in wbt sys-

tems. International Journal on E-Learning (IJEL),

4(4):445–461.

Homma, H. and Aoki, Y. (2003). Creation of wbt server

on digital signal processing. In Proceedings of 4th

International Conference on Information Technology

Based Higher Education and Training. Marrakech,

Morocco.

Kawamura, T., Motomura, S., Nakatani, R., and Sugahara,

K. (2005a). Multi agent-based approach for asyn-

chronous web-based training system. In Proceedings

of International Conference on Integration of Knowl-

edge Intensive Multi-Agent Systems, pages 537–542.

Waltham, Massachusetts, USA.

Kawamura, T., Motomura, S., and Sugahara, K. (2005b).

Implementation of a logic-based multi agent frame-

work on java environment. In Proceedings of Interna-

tional Conference on Integration of Knowledge Inten-

sive Multi-Agent Systems, pages 486–491. Waltham,

Massachusetts, USA.

Kawamura, T. and Sugahara, K. (2005). A mobile

agent-based p2p e-learning system. IPSJ Journal,

46(1):222–225.

Kollhof, J.-K. (2004). jsolait –trac–. http://jsolait.net/.

Maxima Project (2000). Maxima - a gpl cas based on doe-

macsyma. http://maxima.sourceforge.net/.

Motomura, S., Kawamura, T., and Sugahara, K. (2006a). A

logic-based mobile agent framework for web applica-

tions. In Proceedings of the 2nd International Confer-

ence on Web Information Systems and Technologies,

pages 121–126. Setubal, Portugal.

Motomura, S., Kawamura, T., and Sugahara, K. (2006b).

Logic-based mobile agent framework with a concept

of “ﬁeld”. IPSJ Journal, 47(4):1230–1238.

Motomura, S., Nakatani, R., Kawamura, T., and Sugahara,

K. (2006c). Distributed e-learning system using p2p

technology. In Proceedings of the 2nd International

Conference on Web Information Systems and Tech-

nologies, pages 250–255. Setubal, Portugal.

Mozilla Corporation (2005). Rediscover the web.

http://www.mozilla.com/ﬁrefox/.

Napster (2003). napster. http://www.napster.com/.

Nishita, T., Kondo, K., Ohno, Y., Takai, Y., Miura, K. T.,

Dobashi, Y., Ishiuchi, S., Takahashi, T., Kimura, A.,

and Miyai, A. (2002). Development of a web based

training system and courseware for advanced com-

puter graphics courses enhanced by interactive java

applets. In Proceedings of International Conference

on Geometry and Graphics, volume 2, pages 123–

128.

Ratnasamy, S., Francis, P., Handley, M., Karp, R., and

Shenker, S. (2001). A scalable content-addressable

network. In Proceedings of the 2001 conference on

Applications, technologies, architectures, and proto-

cols for computer communications, pages 161–172.

ACM Press.

Winer, D. (1998). Xml-rpc speciﬁcation.

http://www.xmlrpc.com/spec.

USER AND APPLICATION INTERFACE FOR A DISTRIBUTED E-LEARNING SYSTEM

363