DESIGN OF PROACTIVE SCENARIOS AND RULES FOR

ENHANCED E-LEARNING

Sandro Reis, Denis Shirnin and Denis Zampunieris

Computer Science and Communication Research Unit, University of Luxembourg

6, rue Richard Coudenhove-Kalergi, L-1359, Luxembourg City, Luxembourg

Keywords:

Technology Enhanced Learning, Intelligent Tutoring Systems, Proactive System.

Abstract:

We show in this position paper how we designed Proactive Scenarios for an automatic and enhanced manage-

ment of the online assignments on Moodle

for both student and teacher users, through their implementation

with Proactive Rules to be run on top of our prototype Proactive Engine developed for this LMS. According

to the diversity of issues that arise from the users activity on LMS, Proactive Scenarios are of two main types,

which differ in their main goals, features and complexity. Meta Scenarios are devoted to capture major events

of interest and to trigger off the dedicated Target Scenarios, which will undertake the appropriate actions.

These Proactive Scenarios will thus take care of speciﬁcally predeﬁned situations such as notiﬁcations, re-

minders, problem prevention, user guiding etc. In our opinion, LMS supplemented by such capabilities could

provide a boosted effect on the students’ learning process as it takes an individual approach for each user and

therefore could be characterized as a type of intelligent tutoring system. However, in order to sustain or mod-

ify the direction of our research activity, we now consider to undertake empirical studies on real-life online

courses using the Enhanced E-learning Platform, which runs our Proactive Scenarios.

1 INTRODUCTION

Learning Management Systems (LMS) or e-learning

platforms are dedicated software tools intended to

offer a virtual educational and/or training environ-

ment online. Currently available types of LMS are

fundamentally limited tools. Indeed, these systems

have been designed by adapting existing interactive

web-based technologies to deliver learning content on

user’s request, limiting the added value of the LMS to

the users own action and not to the needs of the learn-

ing process.

Proactive Systems, as deﬁned in (Tennenhouse,

2000), adhere to two premises: working on behalf

of, or pro, the user, and acting on their own initia-

tive, without the users explicit command. Proactive

behaviours are intended to cause changes, rather than

just to react to changes.

In (Zampunieris, 2008), we introduced a new

kind of LMS: Proactive LMS, designed to improve

the users online (inter)-actions by providing pro-

grammable, automatic and continuous analysis of the

users behaviours, augmented with appropriate actions

initiated by the LMS itself.

Our Proactive LMS is theoretically able to auto-

matically and continuously help and take care of e-

learners with respect to previously deﬁned procedures

called Proactive Scenarios. Thus, our system is ca-

pable of detecting an “anomalous” behaviour of e-

learner and to communicate the details to concerned

e-teacher; or, the system can check automatically the

awaited behaviours of e-learners, and react if these

actions did not take place. According to particular

contextual characteristics of user’s actions, Proactive

System is able to determine the necessary response

for identiﬁed conditions and to launch the appropriate

Target Scenarios. As an example, the system could

trigger the user’s collaborative attribute of his/her on-

line activity by suggestion of using the speciﬁc fo-

rums and chat-rooms. In (Coronado and Zampunieris,

2010) we reported the statistical analysis of studies

we conducted in a blended learning environment at

the bachelor level. The idea consisted of comparing

a study-group and a control-group of students in the

same course with respect to their intermediate and ﬁ-

nal results. Students of the study-group were continu-

ously triggered by hand-made online messages to in-

cite them to participate to the lectures and interact via

the LMS. Thus we reported that continuous proactiv-

ity supported by the LMS, has direct and positive im-

pact on the students’ learning process.

In this position paper, we show how we designed

253

Reis S., Shirnin D. and Zampunieris D..

DESIGN OF PROACTIVE SCENARIOS AND RULES FOR ENHANCED E-LEARNING.

DOI: 10.5220/0003956302530258

In Proceedings of the 4th International Conference on Computer Supported Education (CSEDU-2012), pages 253-258

ISBN: 978-989-8565-06-8

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

various Proactive Scenarios for an automatic and en-

hanced management of the online assignments sub-

system on Moodle

, a free and open source LMS (see

http://moodle.org), for both student and teacher users.

We explain as well how scenarios are implemented

by means of Proactive Rules and how they are exe-

cuted through the Rules Engine. Finally, we describe

our viewpoints about experimental part of the project

where we expect to collect the feedback in order to

measure the efﬁciency of the Proactive Scenarios and

thus, to enhance its potential, and to validate the ap-

plied research approaches.

2 INTELLIGENT TUTORING

SYSTEMS

In the variety of most popular educational theories

and learning methods, tutoring approach takes a sta-

ble ground and moreover inspires certain researchers

to take the concept for further development and im-

plementations.

The statement that learning process is more effec-

tive and it has a great potential in one-to-one way of

learning was made by Bloom during his research on

adapting teaching. His study discloses that the stu-

dents tutored by master were more successful in their

results with a probability of 98% against the students

with instructional type of classroom teaching (Bloom,

1984). The results of this study have opened a variety

of directions where this model could be taken for its

potential application into another type of the related

ﬁelds. Thus, the idea of computer assisted instruc-

tional programs has already been on the horizon since

1960 (Larkin and Chabay, 1992). However it still

needed the further contribution of research efforts.

Jaime Carbonell has made the signiﬁcant change in

early 1970s when in his Ph.D. thesis he has adopted

the human tutor model into the ﬁrst intelligent tu-

toring program SCHOLAR (Carbonell, 1970a; Car-

bonell, 1970b). The goal of implementation of such

analogy related to human tutor type of learning was

to sustain the reasoning activity of a student basing on

his or her behaviour. In the later years more and more

studies about theories of learning have accentuated

the importance of feedback and practice (Kirschner

et al., 2006). Thus, the research in the ﬁeld of intel-

ligent tutoring systems through decades has emerged

to the dimension where computer science opened the

doors to the advantage of joined research efforts that

have been built together with the collaborative re-

search in cognitive science (Lesgold et al., 1988).

2.1 E-learning Platforms

Learning Management System or LMS was one of

the products, which emerged as the deviation from

the concept of e-learning and computer assisted in-

structional programs. Fundamentally it represents

an online environment that handles different sides of

blended learning such as administrative management

and organisation of virtual courses, different learning

activities, materials etc. The main goal of e-learning

platforms is to track the student’s process of learning

by facilitating the management of various academic

activities.

Taking into consideration all of the advantages of

Learning Management Systems, it could be noticed

however that the LMS misses the essence of proac-

tive type of behaviour, which could in our opinion sig-

niﬁcantly increase the outcomes of students’ learning

process.

2.2 Proactivity as the Tool for Enhanced

E-learning

The notion of Proactive Systems has been intro-

duced by Tennenhouse. He described its functional-

ity as the mechanism, which interacts with the world

around it, using sensors and actuators (Tennenhouse,

2000). The sensors’ implementation serves as the

perception-center of the system, which is able to cap-

ture and observe an event of interest and perform the

appropriate actions on its own initiative. The original

idea has pushed researchers in computer science to

take this approach to another level for further devel-

opment. Thus, the potential of proactive systems has

found a stable ground in the ﬁeld of Learning Man-

agement Systems (LMS).

Proactive or Context-aware Learning Manage-

ment Systems (PLMS) basing on users’ activity

and its data analysis are capable of acting semi-

autonomously or without explicit instructions from

the user (Salovaara and Oulasvirta, 2004; Zam-

punieris, 2006; Zampunieris, 2008). Due to advan-

tages of LMS such as integration with other software

solutions, we take it into another level where the or-

dinary e-learning platform will be provided with the

proactive type of behaviour (PLMS).

LMS supplemented by such capabilities in our

opinion could provide a boosted effect on students’

learning process. Considering, on the one hand that

this process is a result of collaborative work, it pro-

vides the best environment for our Proactive System.

On the other hand, PLMS takes an individual ap-

proach for each user (student, teacher, system admin-

istrator) and therefore could be characterised as a type

of tutoring system.

CSEDU2012-4thInternationalConferenceonComputerSupportedEducation

254

2.3 Proactive System Integration in

LMS

The motivation surrounding the enhancement of the

learning progress of the students is not new and with

the new digital age all old ideas change their shapes

into the direction of computerisation. Thus, again,

the question how to improve students’ academic re-

sults has changed to how to help the students in

their learning process, by means of new technologies.

Certainly we get the good results from implement-

ing these computer-based technologies (Regian et al.,

1996). On the other hand what do we get if we com-

bine different types of new computer-based tools with

the objective to enhance its productive ratio, for ex-

ample, in the form of better academic results for the

students or better management of the learning con-

tent and tasks? Combination of proactivity with the

e-learning platform could help or assist user while he

or she performs certain tasks in LMS environment. In

its totality, PLMS represents a groupware tool, which

aims to boost the online academic activity of a student

as well as the effectiveness of their learning process

by providing a variety of Proactive Scenarios, which

potentially covers all possible situations arising from

different activities and tasks.

2.4 Prototype of PLMS on Moodle

This project has a premise that the Proactive Engine

will be embedded into Moodle

, the e-learning plat-

form currently used at the University of Luxembourg.

Such approach allows the research team to focus on

the design and implementation of Proactive Engine,

using existing LMS as a framework.

Figure 1: System architecture.

As shown in the Figure 1, the core of the Proactive

System is the Rules Engine, which is responsible for

the control ﬂow of the rules execution where the rules

represent the implementation of the Proactive Scenar-

ios’ logic. Each rule is implemented in a separate

Java class. The idea of the Rules Engine is to add

the proactive behaviour to the LMS, but not to change

its code or to restrict its execution. Several parame-

ters were introduced in (Zampunieris, 2006) in order

to conﬁgure the PLMS system in a way that the main

core functions of LMS and Proactive System don’t

overload the host computer. Hereafter are the exam-

ples of mentioned parameters: (F) - Time frequency

of its activation periods; (N) - The (maximum) num-

ber of rules it runs in an activation period. We also

introduced a third parameter to ensure that our sys-

tem doesn’t interfere with the LMS main process: (P)

- The (minimum) time Proactive System pauses be-

tween two activation periods.

The Rules Engine acts as it was deﬁned in (Zam-

punieris, 2008), where it is responsible for storing the

set of rules. It also stores a second list of rules that

represents the rules generated in the current activation

period. This second list will be added at the end to the

remaining set of rules (in case the system parameter

(N) has been reached), in order to be executed in the

next activation period.

The interaction with Moodle

is done via its

database, where Proactive System checks changes of

its state, relevant to the scenarios’ logic. To that ef-

fect, we developed an abstract database wrapper, with

two implementations: MySQL (with the textual SQL

queries needed to access Moodle’s data) and MySQL

with a cache on top of the ﬁrst one (applying the

Proxy Design Pattern). We ensure the persistence of

the Rules Engine by using the Hibernate

framework

to store the rules’ queue, the generated messages and

some system statistics, on a speciﬁc database schema.

As for interacting with the user, the system sends

emails and/or messages embedded into the LMS sys-

tem, depending on whether the user is online or of-

ﬂine. We are currently developing the Moodle

add-

on, which permits the user to interact with our sys-

tem’s messages. We are working as well on the ad-

ministrator’s interface, which will include such menu

functions as starting/stopping the engine, and chang-

ing the system parameters.

3 PROACTIVE SCENARIOS

Proactive System is the goal-oriented mechanism,

which entails a set of scenarios with an objective to

provide a help to the user or the tutor according to

their activity on LMS. Scenarios have different areas

of application; it typiﬁes the nature and its main oper-

DESIGNOFPROACTIVESCENARIOSANDRULESFORENHANCEDE-LEARNING

255

ational directions. According to the diversity of issues

that arise from the user’s activity on LMS, the scenar-

ios may differ in its features, essence, and complexity.

Two types deﬁne the category of Proactive Sce-

narios: type #1 are the Meta Scenarios, and type #2

are the Target Scenarios.

3.1 Type #1: Meta Scenarios

The goal of scenario of type #1 is to provide the sys-

tem with feature of the perception-center. That is, to

capture an event of interest and to undertake the ap-

propriate actions.

In order to activate the speciﬁc scenario, which

will correspond to the actual situation of the user’s

activity, the system needs to be aware about the cur-

rent state of the LMS database. As the Target Sce-

narios have not the capability to detect any changes

on LMS but only perform the speciﬁc job, this role is

attributed to the so-called Meta Scenarios. The main

functionality of this type of scenario is to be context

aware continuous never-ending rule. As soon as the

Meta Scenario detects the corresponding event on the

LMS, it activates the Target Scenarios, which will do

the predeﬁned actions or in different words the Meta

Scenario will delegate the speciﬁc job to the appropri-

ate scenarios.

Basic type of implementation of Meta Scenarios

is the system environment of LMS. It means that this

type of scenario will provide our system with the in-

teractions between user and Proactive System as well

as Proactive System and LMS database. The actions

of Meta Scenario are characterised as inward related.

However, the effect of these actions is mostly oriented

on the outer user’s environment.

3.2 Type #2: Target Scenarios

The goal of scenario of type #2 is to provide the mul-

tiple target responses to each detected by Meta Sce-

nario event or non-event.

In metaphorical perspective, the scenarios of type

#2 are to be the hands of the Proactive System. They

are responsible for the single target actions that have

been initiated by the Meta Scenarios. The type #2

takes care only of speciﬁcally predeﬁned situations

such as Notiﬁcations that aim to inform a user about

an event, Reminders, Problem prevention, User guid-

ing etc. Taking into the account that all scenarios are

nothing else as the set of rules, after having performed

its individual job each rule becomes dismissed. This

is the radical difference between two types of scenar-

ios. When Meta Scenario is deﬁned to be the never-

ending rule, the Target Scenario simply dies after each

completed task. It permits to optimize our system in

terms of memory usage.

In the similar perspective as for the Meta Scenar-

ios, the Target Scenarios have their own areas of ap-

plication. The basic characteristic of the rules em-

ployment of that type is its outward direction of the

actions. In our case the focus is deﬁned by three

different orientations: system administrator environ-

ment, e-teacher environment, and e-student environ-

ment. Thus, while creating new scenarios and rules

we try to maximize the accuracy of the deﬁned ac-

tions’ outcome of the Proactive System and to better

respond on the detected need arisen from the users ac-

tions. In order to do so, we have to pay an attention on

the cognitive aspect of user’s intentions, objectives,

and actions.

3.3 Use of Joined Approaches

One of the main objectives of the study is to build the

scientiﬁc evidence for the outcomes of our research

activity. Thus, we found the combination of two dif-

ferent domains, computer science and cognitive sci-

ence, beneﬁcial.

While working on the Proactive Scenarios we

have to analyse the perspective of different cogni-

tive approaches in the users’ behaviour or, in other

words the speciﬁc context related activity in order to

provide the response from our Proactive System with

the accurate actions based on the user-oriented meth-

ods of cognitive expertise. Different cognitive the-

ories, which are outlined below, are applied during

this process. We believe that the results issued from

the joined approaches may provide us with the objec-

tive and accurately grounded scientiﬁc evidence for

the further research efforts in this ﬁeld.

The following brief description of the concept the-

ories used in our research, highlights the main orienta-

tions that we consider currently in the process of plan-

ning and creating the Proactive Scenarios and rules.

Cognitive Approach and Behavioural Science

provide us with possibility to study and to display the

user’s behaviour while working online, to help us to

build the schema of possible users actions in speciﬁc

situations and accordingly to implement these aspects

into Proactive Scenarios (Burnes, 2005; Gao et al.,

2002).

Theory of Socially Shared Cognition is linked

to the theory of Computer-user Interactions, and

will display the aspects of computer-mediated inter-

actions, synchronous or asynchronous. By taking

the examples from social interactions and studying

them we could ﬁnd the equivalent type of application

in human-computer interactions (Siler, 2009; Wrede

CSEDU2012-4thInternationalConferenceonComputerSupportedEducation

256

et al., 2010; Yeh et al., 2007).

Activity Theory is partially linked to the theory of

Socially Shared Cognition and will help us to display

the aspects of how learning takes place basing on the

Higher Mental Functions Theory of Vygotsky (Nardi,

1996; Vygotsky, 1981).

Theory of User’s Identity gives us the picture of

the average user where we deﬁne his/her fundamen-

tal behavioural patterns and apply them into Proactive

Scenarios (Rowe, 2010; Zimmerman, 1998).

3.4 Map Representation of Scenarios

Connections

Figure 2 provides the visual representation of all sce-

narios connections in the form of a decision tree. It

shows the process of how one Meta Scenario accord-

ing to the captured data from the inward system en-

vironment or outer user environment launches several

Target Scenarios. The hierarchy of Proactive Scenar-

ios is indicated by the speciﬁc colour, which distin-

guishes and regroups the similar layers in one cate-

gory.

Thus, all Meta Scenarios abbreviated as MTA bas-

ing on speciﬁc parameters deploy the appropriate path

of actions though the set of Target Scenarios. As soon

as the subjected task is accomplished, the proactive

process jumps back to the level of Meta Scenarios

where it continues to look for the new data.

Figure 2: Generic representation of scenarios connections.

4 FUTURE EFFORTS

While creating the new unveriﬁed yet technology,

there is a need for the testing and validation of the

ideas, theories, and potential results derived from the

archetypal phase of the research project. Thus, in or-

der to sustain or modify the direction of our research

activity, we consider to undertake the empirical study

and the analysis of the data issued from the upcoming

experiments.

The sessions of experiments will take place be-

tween February and June of 2012 at the University of

Luxembourg. The participants of the experiment are

the students enrolled in the bachelor program at the

faculty of Computer Science and Communications.

The goals of the experiments are divided into two

modules. The ﬁrst aims to improve our Proactive

System’s functionality with the particular emphasis

on Queue Manager, User Interface, Messaging Sys-

tem, and elaboration of Proactive Scenarios. For the

second module the objective is to enhance the stu-

dents’ success level in online virtual academic ac-

tivity as well as to boost their learning process. In

order to do so, we divided main goal into the sub-

categories of speciﬁc objectives such as students’ e-

learning practice, their cooperative and collaborative

learning skills, learning competences, learning expe-

rience, and learning performance.

Thus, at the present moment we develop the mea-

surement tools, which allow us to test each cate-

gory and subcategory of the deﬁned goals of the ex-

periments. Such measurement tools aim to collect

the feedback from the users whether automatically

through the statistics reports of the system or man-

ually through online surveys, interviews, question-

naires, and live discussions. For some categories of

experiment we will use the similar techniques as in

(Coronado and Zampunieris, 2010).

We assume that this experiment will help us to

detect the potential research gaps and to enhance the

probability of valid and constructive outcomes.

5 CONCLUSIONS

The position paper describes the main concepts of the

Proactive System. Implemented into the Moodle

, it

aims to enhance the capabilities of LMS by enrich-

ing its main functions with the proactive type of be-

haviour. As reported in previous experiments, the fea-

ture of proactivity has positive reﬂection on e-learning

experience of the students.

According to the diversity of issues that arise from

the users activity on LMS, Proactive Scenarios are of

two main types, which differ in their main goals, fea-

tures and complexity. Meta Scenarios are devoted to

capture major events of interest and to trigger off the

dedicated Target Scenarios, which will undertake the

appropriate actions. Thus, Proactive Scenarios will

DESIGNOFPROACTIVESCENARIOSANDRULESFORENHANCEDE-LEARNING

257

take care of speciﬁcally predeﬁned situations such as

Notiﬁcations, Reminders, Problem prevention, User

guiding etc. In our opinion, LMS supplemented by

such capabilities could provide a boosted effect on

the students’ learning process as the Proactive LMS

takes an individual approach for each user and there-

fore could be characterised as a type of intelligent tu-

toring system. However, in order to sustain or mod-

ify the direction of our research activity, we now con-

sider to undertake empirical studies on real-life on-

line courses using the enhanced e-learning platform,

which runs our Proactive Scenarios. We expect to col-

lect the feedback in order to measure the efﬁciency of

the Proactive Scenarios and thus, to enhance its poten-

tial, and to validate the applied research approaches.

REFERENCES

Bloom, B. S. (1984). The 2 sigma problem: The search for

methods of group instruction as effective as one-to-

one tutoring. Educational Researcher, 13:4–16.

Burnes, B. (2005). Complexity theories and organizational

change. International Journal of Management Re-

views, 7(2):73–90.

Carbonell, J. R. (1970a). AI in CAI: an artiﬁcial-

intelligence approach to computer-assisted instruc-

tion. IEEE Transactions on Man-Machine Systems,

11(4):190–202.

Carbonell, J. R. (1970b). Mixed-initiative man-computer in-

structional dialogues. Final report. MIT Press, Cam-

bridge, MA.

Coronado, S. and Zampunieris, D. (2010). Continuous

Proactivity in Learning Management Systems. In

IEEE EDUCON 2010 Conference, pages 199–204.

Gao, F., Li, M., and Nakamori, Y. (2002). Critical systems

thinking as a way to manage knowledge. Systems Re-

search and Behavioral Science, 20(1):3–19.

Kirschner, P. A., Sweller, J., and Clark, R. E. (2006).

Why Minimal Guidance During Instruction Does Not

Work: An Analysis of the Failure of Constructivist,

Discovery, Problem-Based, Experiential, and Inquiry-

Based Teaching. Educational Psychologist, 41(2):75–

86.

Larkin, J. H. and Chabay, R. W. (1992). Computer-Assisted

Instruction and intelligent Tutoring Systems: Shared

Goals and Complementary Approaches. Lawrence

Erlbaum Associates, Inc., Hillsdale, NJ.

Lesgold, A., Lajoie, S., Bunzo, M., and Eggan, G. (1988).

SHERLOCK: A Coached Practice Environment for an

Electronics Troubleshooting Job. Defense Technical

Information Center, Pittsburgh, PA.

Nardi, B. A. (1996). Context and consciousness: Activity

theory and human-computer interactions. MIT Press.

Regian, J. W., Seidel, R. J., Schuler, J., and Radtke,

P. (1996). Functional Area Analysis of Intelligent

Computer-Assisted Instruction. Training and Person-

nel Systems Science and Technology Evaluation and

Management Committee, USA.

Rowe, M. (2010). The credibility of digital identity infor-

mation on the social web: a user study. In Proceedings

of the 4th workshop on Information credibility, pages

35–42. ACM.

Salovaara, A. and Oulasvirta, A. (2004). Six modes of

proactive resource management: a user-centric typol-

ogy for proactive behaviors. In Proceedings of the

third Nordic conference on Human-computer interac-

tion, pages 57–60. ACM.

Siler, S. A. (2009). Learning, Interactional, and Mo-

tivational Outcomes in One-to-One Synchronous

Computer-mediated versus Face-to-Face Tutoring. In-

ternational Journal of Artiﬁcial Intelligence and Edu-

cation, 19(1).

Tennenhouse, D. (2000). Proactive computing. Communi-

cations of the ACM, 43(5):43–50.

Vygotsky, L. S. (1981). The genesis of higher mental func-

tions. In Wertsch, J. V., editor, The concept of activ-

ity in Soviet psychology, pages 144–188. Sharpe, Ar-

monk, NY.

Wrede, B., Kopp, S., Rohlﬁng, K., Lohse, M., and Muhl, C.

(2010). Appropriate feedback in asymmetric interac-

tions. Journal of Pragmatics, 42(9):2369–2384.

Yeh, S.-W., Lo, J.-J., Huang, J.-J., and Fan, Z.-Y. (2007).

A Synchronous Scaffolding Environment for Col-

laborative Technical Writing. In Technologies for

E-Learning and Digital Entertainment. Edutainment

2007, volume 4469, pages 829–840, Berlin Heidel-

berg. Springer.

Zampunieris, D. (2006). Implementation of a Proactive

Learning Management System. In E-Learn World

Conference on E-Learning in Corporate, Govern-

ment, Healthcare and Higher Education, pages 3145–

3151, Hawaii.

Zampunieris, D. (2008). Implementation of Efﬁcient

Proactive Computing Using Lazy Evaluation in a

Learning Management System. International Jour-

nal of Web-Based Learning and Teaching Technolo-

gies, 3(1):103–109.

Zimmerman, D. H. (1998). Identity, context and interaction.

In Antaki, C. and Widdicombe, S., editors, Identities

in Talk, pages 87–106. Sage, London.

CSEDU2012-4thInternationalConferenceonComputerSupportedEducation

258