An Intelligent Learning Support System

Mariia Gavriushenko, Oleksiy Khriyenko and Ari Tuhkala

Faculty of Information Technology, University of Jyv¨askyl¨a, Jyv¨askyl¨a, Finland

Keywords:

Intelligent Learning System, Adaptive and Personalized Education, Career Development.

Abstract:

Fast-growing technologies are shaping many aspects of societies. Educational systems, in general, are still

rather traditional: learner applies for school or university, chooses the subject, takes the courses, and ﬁnally

graduates. The problem is that labor markets are constantly changing and the needed professional skills

might not match with the curriculum of the educational program. It might be that it is not even possible to

learn a combination of desired skills within one educational organization. For example, there are only a few

universities that can provide high-quality teaching in several different areas. Therefore, learners may have to

study speciﬁc modules and units somewhere else, for example, in massive open online courses. A person, who

is learning some particular content from outside of the university, could have some knowledge gaps which

should be recognized. We argue that it is possible to respond to these challenges with adaptive, intelligent, and

personalized learning systems that utilize data analytics, machine learning, and Semantic Web technologies.

In this paper, we propose a model for an Intelligent Learning Support System that guides learner during the

whole lifecycle using semantic annotation methodology. Semantic annotation of learning materials is done not

only on the course level but also at the content level to perform semantic reasoning about the possible learning

gaps. Based on this reasoning, the system can recommend extensive learning material.

1 INTRODUCTION

Many young people face difﬁculties when trying to

decide what kind of education to apply. Once the ap-

proval to the wanted education is achieved, many de-

cisions must be made: which subjects to choose, what

courses to put to the personal study plan, what courses

are relevant in future job markets, or if the ofﬁcial cur-

riculum should be supported with supplemental edu-

cation from the private sector. These decisions might

have life-long effects, but often the resources for mak-

ing the right choices are limited. People should rec-

ognize their competitive advantages and put them in

the perspective of global trends, technological devel-

opment, and labor markets. Not only young people

have such need, but many graduates and experienced

workers also face the situation of economic change,

and suddenly they have to adjust their competence by

improving and extending their skills.

In welfare countries, ﬂexible education system of-

fers quite a many options to choose. The possibili-

ties are so widespread that it is not easy to make the

right choice without a comprehensive analysis of per-

son’s capabilities, ambitions, and values against the

future labor markets. However, only a small amount

of people have an actual freedom to choose their fu-

ture profession. Usually, people have to be competi-

tive enough to survive, especially in the times of cri-

sis, and those who have the possibility, are changing

their location to get a better education for themselves.

On the contrary, it has become possible to get the ed-

ucation from another side of the world, especially in

higher education. Massive open online courses and

learning platforms like Udemy, Khan Academy, and

Coursera are helping people to get desired skills (e.g.,

programming, developer skills, and data analytics)

without leaving home. However, some courses have

pre-requirements, so that it would be quite difﬁcult to

attend courses without ﬁlling the existing knowledge

gaps by studying an additional learning material.

Educational systems are structured around curric-

ula that aim to supply people with the deﬁned com-

petencies. The problem is that employers are more

and more ﬁnding people with various skills and back-

grounds, which may not be in line with the cur-

riculum. Therefore, people should be able to create

speciﬁc curriculum and be educated following per-

sonal learning trajectory. To meet this demand, top-

ics around different learning support systems have be-

come popular. For example, the Open University of

Hong Kong has developed a system that instantly re-

sponds to inquiries about career development, pro-

Gavriushenko, M., Khriyenko, O. and Tuhkala, A.

An Intelligent Lear ning Support System.

DOI: 10.5220/0006252102170225

In Proceedings of the 9th International Conference on Computer Supported Education (CSEDU 2017) - Volume 1, pages 217-225

ISBN: 978-989-758-239-4

217

gram and course choices, study plans, and gradua-

tion checks (Leung et al., 2010). While these systems

have potential, the problem is that they should be con-

nected with various educational and career planning

services. We can already verify this issue by look-

ing how many different job portals there exists, both

nationally and globally.

We argue that there is a need for an Intelligent

Learning Support System (ILSS), which will be based

on current research information about adaptive, per-

sonalized, and intelligent tools. We begin by review-

ing the current body of knowledge about these topics.

Next, we explain the whole ecosystem of ILSS and

the components it includes. Finally, we discuss impli-

cations that must be considered when implementing

this kind of system. This proposition paper is cur-

rent because the development of adaptive, personal-

ized, and intelligent systems for education are still in

the premature phase in Finland, where technological

infrastructure is Finland under reconstruction: The

government’s aim is to standardize and integrate the

social and health care systems, including the career

and employment services. Students, job seekers, em-

ployment ofﬁcials, and company headhunters would

beneﬁt from a centralized system that aims to make

educational interests and employment supply meet.

2 LITERATURE REVIEW

Each person has different knowledge background,

ability to memorize, learning speed, motivation, and

preferences. Furthermore, people have different

health conditions and habits when and how to learn.

The problem with educational applications is that they

are built for a speciﬁc group of learners and not indi-

vidually. Therefore, development of learning environ-

ments with individualized learning mechanisms is an

important research topic. Researchers are interested

in developing learning environments, which could

adaptively provide learning path (Chen, 2008), but

the challenge is the development of advanced learn-

ing applications that include intelligence and adaptiv-

ity (Brusilovsky and Peylo, 2003).

Intelligent learning systems are related to adap-

tive learning systems and have an intersection

(Brusilovsky and Peylo, 2003). Intelligent systems

are ”systems which use techniques from Artiﬁcial In-

telligence to provide better support and feedback for

learners” (Brusilovsky and Peylo, 2003) and adaptive

systems are ”systems which attempt to be different

for each learner or group of learners” according to in-

formation about learners received during their actions

(Brusilovsky and Peylo, 2003). Personalized systems

present a speciﬁc type of general adaptive systems

(Garc´ıa-Barrios et al., 2005) and they mean adapta-

tion towards a speciﬁc learner.

Computer based training or computer aided learn-

ing does not use a model of the learner’s knowl-

edge and does not take into account personality of the

learner, but it just uses traditional instructional meth-

ods (Phobun and Vicheanpanya, 2010). The com-

bination of adaptation, intelligence and personaliza-

tion could make this better. The interaction with the

learner would not only report the correct or incorrect

answers, but also explanations of the error cause, and

recommendations for study material.

The development of adaptive, intelligent, and per-

sonalized learning systems is important, especially

for children with disabilities, because they require

more personalization during their learning process.

These techniques should allow them to learn at their

pace and help to keep their motivation (Gavriushenko

et al., 2016).

2.1 Personalized Learning Systems

Many authors suggest in their work different tools

for learning systems, which could assist in learning

and make this process more personalized. For ex-

ample, in the work (Bendakir and A¨ımeur, 2006)

authors presented a course recommendation system

which analyzes past behavior of learners concern-

ing their choices. Authors used data mining associa-

tion rules (user ratings), and this classiﬁcation makes

possible to build a decision tree that classiﬁes each

learner proﬁles. The system recommends learning

path by checking which courses are followed simul-

taneously. Unfortunately, proposed system does not

consider learners’ background and course availabil-

ity. Also, researchers (Werghi and Kamoun, 2009)

have presented a decision support system for learner

advising which provides an automated program plan-

ning and scheduling service that ﬁts best their pro-

ﬁles while meeting academic requirements. The sys-

tem was based on decision tree algorithm. The work

(Jeong et al., 2012) presented a personalized learn-

ing course planner with decision support using user

proﬁle. This system allows the learner to select the

learning course they desire taking into account pre-

vious learning information of learner. Authors used

organization algorithm containing the decision ma-

trix. The proposed system improved learning effec-

tiveness and especially learners’ satisfactory accord-

ing to the questionnaire. Also, in a paper (Jyothi et al.,

2012) was presented a recommender system which

assists the instructor in building learning path, more

speciﬁcally in identifying the groups of learners who

CSEDU 2017 - 9th International Conference on Computer Supported Education

218

have similar learning styles and then providing rec-

ommendations to these learners. Authors used Felder-

Silverman learning style model.

2.2 Intelligent and Adaptive Learning

Systems

Some research papers were concentrated on develop-

ing intelligent and adaptive applications for learning

which could fully or partially reduce the involvement

of human advisers. For example, in the work (Hen-

derson and Goodridge, 2015) authors proposed a sys-

tem which gives accurate advice only to those pur-

suing ”special degrees” or programs which follow a

clear cut path of courses. This system is an intelli-

gent Web-based application for the handling of gen-

eral advising cases. Authors used Semantic Web tech-

nologies for designing this system. Also, in the work

(Wen and McGreal, 2007) system allows learners to

add preferences of the specialization to their proﬁle

and then recommend courses based on their prefer-

ences. The proposed system uses a multiple intel-

ligent agents approach and ontology-driven method-

ology to tackle a dynamic and complex individual-

ized study planning problem. In other research, (Nur-

janah, 2016) authors proposed recommender system

in adaptive learning which recommends learning ma-

terials for learners. This system combines content-

based ﬁltering and collaborative ﬁltering approach

which is based on the similarity between learners and

learners’ competencies. The result of the proposed

new technique of recommendation showed that it per-

forms well. In paper (Myneni et al., 2013) authors

presented an interactive and intelligent learning sys-

tem for physics education. This environment helps

learners master physics concepts in the context of pul-

leys. System guides learners in problem-solving. Au-

thors used the Bayesian network for pulley setup. In

the work (Dolenc and Aberˇsek, 2015) authors demon-

strate the design and evaluation of intelligent tutor-

ing system based on cognitive characteristics of the

learner. This model is based on a system for collect-

ing metadata and variables that are vital for the teach-

ing process. The important thing in this research is

that system can be adapted to the different level of

learner’s knowledge and understanding subject mat-

ters. Authors in paper (Canales et al., 2007) suggest

an adaptive and intelligent Web-based education sys-

tems that take into account the individual learning re-

quirements. They proposed three modules for that:

an Authoring tool, a Semantic Web-based Evaluation,

and Cognitive Maps-based Learner Model. Proposed

approach is focused on reusability, accessibility, dura-

bility, and interoperability of the learning content.

There are many intelligent tools developed for

learning systems which are effective and shows good

results. However, there are not much tools or systems

which could help learner during his whole lifecycle.

3 ARGUMENTATION FOR THE

ILSS

This section proposes the ecosystem of the Intelligent

Learning Support System (as presented in ﬁgure 1)

and describes the main mechanisms that allow learn-

ers to get learning content recommendations. The

model of the system based on semantic technology.

At ﬁrst, we discuss using Semantic Web in building

ILSS. After, the main blocks of the system are pre-

sented in sub-chapters. The proposed system is now

in the development stage and includes the necessary

functionality description required.

3.1 Using Semantic Web in Building

ILSS

Developing adaptive, intelligent, and personalized

systems require comprehensive knowledge in the do-

main and ﬂexible mechanism for further extension of

it as well as the automated mechanism for contextual

information retrieval. It will require representation of

knowledge, which is widely distributed among dif-

ferent organizations, freely available, and presented

in machine readable form. The system should adopt

beneﬁts of Semantic Web and linked data technolo-

gies, as well as, be facilitated by text mining and nat-

ural language processing techniques.

There has been many studies proposed for

the intelligent study advising utilizing Semantic

Web technologies (Henderson and Goodridge, 2015;

Gavriushenko et al., 2015; Ranganathan and Brown,

2007; Nguyen et al., 2008; Jovanovi´c et al., 2007).

Authors in (Dicheva et al., 2009; Jovanovic et al.,

2007) used ontologies and Semantic Web in cur-

riculum development and in institutional support and

adaptation. Semantic Web helps formalize learning

object content, shows how to use semantic annota-

tion to interrelate diverse learning artifacts (Jovanovi´c

et al., 2009).

Semantic Web aims at integrating various types

of information into a single structure, where each

semantic data element will meet special syntactic

block. It provides a common framework that al-

lows data to be shared and reused across application,

enterprise, and community boundaries (Berners-Lee

et al., 2001). Semantic Web achieving interoperabil-

An Intelligent Learning Support System

219

Figure 1: General ecosystem of an Intelligent Learning Support System.

ity among various educational systems and uniﬁed au-

thoring support for the creation (Aroyo and Dicheva,

2004). The other branch of the Semantic Web asso-

ciated with a direction close to the ﬁeld of artiﬁcial

intelligence, and it is called the ontological approach.

The ontological approach means a formal repre-

sentations of a set of concepts within a domain and

the relationships between those concepts (knowledge

base). Under the knowledge base, we understand

complex data structure that stores the knowledge do-

main. Typically, the knowledge base is represented as

a graph in which the vertexes are one unit of learn-

ing, that is the simple essence of knowledge (train-

ing course, the term, the deﬁnition of the formula –

depends on an agreement between system’s makers

that they will assume a basic understanding of the

essence). The arcs are seeing as all kinds of links be-

tween training units.

Knowledge bases are used for solving problems

like creation, transformation, and usage of content

while learning. The engineering part includes the cre-

ation of the educational plans and their e-learning re-

sources. Management part problems are solved di-

rectly during the learning process and are focused on

operational adjustment training course depending on

the results of the test control of learners’ intermediate

mastering of the material. Creation of the ontologies

as knowledge bases are very useful for tests creation,

automation of the evaluation, and management of the

training paths based on the results of testing.

To be processed by intelligence (applications, ser-

vices) and be consumed by end-user in a best opti-

mal way, learning content should be adapted to the

digital world of intelligent entities and must be sup-

plied with correspondent semantic annotation. Se-

mantic annotation provides a set of worldwide stan-

dards, which helps in operating with heterogeneous

resources using common syntax and methods (Uren

et al., 2006). Semantic annotation by the rules iden-

tiﬁes concepts and their relationships, and it is meant

for use by machines (Uren et al., 2006). Semantic

annotation was presented and successfully applied in

papers (Jovanovi´c et al., 2009; Jovanovi´c et al., 2007),

where authors presented ontology-based approach to

automate annotation of learning objects and tested it

in integrated learning environment for domain of In-

telligent Information Systems.

Providing the personalized point of view on con-

tent annotation, learner extends a pool of context de-

pendent semantic annotations for the same content.

Later, annotations might be transformed into a set of

different annotations for the same content on the dif-

ferent clusters of the learners as well as to the wide

variety of other contexts. Such clustered semantic an-

notations will enhance matchmaking for new content

retrieving and personalized content recommendation

processes.

CSEDU 2017 - 9th International Conference on Computer Supported Education

220

3.2 Courses and Learning Material

Annotation Module and Annotated

Learning Content mModule

Usually, educational material and content are pre-

sented in the same way for everybody, without tak-

ing into account learner’s existing knowledge, goals

for browsing, preferences, and experience. This is-

sue needs more attention, especially when it comes

to e-learning because the learner is supposed to take

responsibility for one’s studies. The main factors

that should be taken into account are learner’s knowl-

edge background, age, experience, motivations, pro-

fessions and goals (Huang et al., 2007).

The common feature of learning support systems

is that they have some grade range for learners of dif-

ferent learning levels. Usually, there is some pre-test

which can determine learner knowledge and create

simple learning path according to the particular level

of the learner. A simple adaptation algorithm adds

that if the learner makes lots of mistakes, the system

returns the learner to the previous level. Otherwise,

the system allows the learner to proceed. The thresh-

old of right answers can be eighty percent. Unfor-

tunately, twenty percent of wrong answers can mean

that some critical knowledge may be missed, which

can interrupt the process in future. It is important to

keep track what learner does not exactly know and

how this particular knowledge could be improved.

The main idea is to ﬁnd learning gaps which inﬂuence

the learning process and recommend material to help

the learner to fulﬁll those gaps. Same kind of learn-

ing gaps can also occur when the learner is search-

ing for a speciﬁc skill and has to take some additional

courses that are not included in the curriculum. The

targeted courses might have prerequisites, which the

person can not meet at the moment.

From the employer’s point of view, it is not always

possible to recognize the potential of a job seeker

from a list of completed courses or degrees. The

names of different courses do not always say much

about actual content, or even opposite may bring

wrong expectations about skillfulness of a person who

passed them. Depending on the educational organi-

zation, there might be a huge difference between the

content of courses with similar names.

It might be more reasonable to see an approximate

level of particular skills of a person, to see an expec-

tation of the potential to resolve the certain class of

problems. Therefore, all the course descriptions or

meta-information should include some explicit refer-

ence to skills and knowledge that it is aimed to lead

or inﬂuence. Thus, having such skills and knowledge

oriented annotation of courses and learning materi-

als, it is possible to facilitate a better guidance for a

learner.

If we make a straightforward assumption, that the

aim of education is to develop certain knowledge and

skills to be applied in different occupations, learner

chooses education steps (university, program, course,

lecture) for a certain goal (profession, career, occu-

pation). Unfortunately, the learner may not be aware

enough of the required skills or knowledge for this

goal. Therefore, occupations and job descriptions

should include skills- and knowledge-oriented anno-

tation as well.

In this work we use semantic annotations with

the use of ontologies (as presented in ﬁgure 2). On-

tologies were used for storing and modeling informa-

tion about learners, teachers, courses, specializations,

universities, careers, job offers, among other things.

With ontologies it is possible in future to apply meth-

ods of semantic alignment for matching the ontolo-

gies (Heath and Bizer, 2011; Berners-Lee, 2006; Hen-

derson and Goodridge, 2015), if there will be some

other existing ontologies which are related to our ed-

ucational system. Nowadays, many tools exist for the

creation and matching ontologies which have proven

themselves well (Berners-Lee et al., 2001; Shvaiko

and Euzenat, 2013). The implementation process was

done by Prot´eg´e-tool which is the most widely used

and offers a complete development environment.

3.3 Personal Learner Proﬁle Module

The educational system produces a wide amount of

data related to learning, performance, difﬁculties, and

experiences. This information can be stored in a Per-

sonal Learner Proﬁle (PLP), which always support

changes and keeps the information up-to-date. In-

terfaces to proﬁle give access to only authenticated

learners and systems. The knowledge basewould give

anonymous data for pedagogical development pur-

poses. Individual learners would be able to access the

information related to themselves. Information stored

in the personal proﬁle would consist of: study history,

work experience, personality, learning plans, learning

difﬁculties, hobby, and social activities.

The PLP would be connected to the external sys-

tem and services with learner’s proﬁles that connects

different systems, services, and databases, for exam-

ple, the national social and health systems. The ILSS

will be developed to analyze the information from

systems connected with the digital service channel.

The system is based on current research knowledge,

and it will have various kind of pedagogical informa-

tion related to learning theories, practices, environ-

ments, and special needs.

An Intelligent Learning Support System

221

Figure 2: Proposed ontology for an Intelligent Learning Support System.

The PLP contains current skills- and knowledge-

oriented proﬁle of a learner in the form of a vector

in multi-dimensional space of knowledge and skills.

The Career Navigator module helps the learner to

build a desired/aimed space of knowledge and skills,

based on selected occupations/jobs the learner aims.

The Career Navigator is connected to the Job market

place that contains: job and occupation descriptions

published by employers, national needs and require-

ments, and expectations of society. The Career Nav-

igator is facilitated with proactive ”Job Radar” mod-

ule that monitors/analyses changes in the Job market-

place and informs learner about latest trends. The Ca-

reer Navigator tool should match curricula, various

certiﬁcates, job history and other measured skills of

a person with the description of profession and job

position requirements. Depending on the set goal,

the navigator may build a career path to the aimed

point, highlight necessary steps and required missing

skills for it, offer relevant study programs and on-

line courses to extend personal skills and education.

Furthermore, it can show the closest job offers from

the labor market, taking into account latest trends in

economics development on regional and governmen-

tal levels or even worldwide, and offer suitable new

profession obtaining programs supported by the gov-

ernment for the temporally unemployed person.

This module should have a possibility to ﬁnd the

program that is more close to the learner’s goal of

ﬁnding the desired job. Learner proﬁle, career, and

study content should be described with the same an-

notation type.

3.4 Personal Study Planner Module

The PLP module works as an input for a Digital Study

Advisor (DSA), which uses a Personal Study Planner

(PSP) to build corresponding study plan and compose

the most suitable study content for a learner, depend-

ing on the available learning materials. Learning Ma-

terials Space Manager is connected to various online

learning platforms and courses, has access to an open

source learning materials, and allows the learner to

extend the space with courses provided by any univer-

sity or education organization learner is enrolled. The

DSA analyses the gaps (missing parts) between the

current skills-/knowledge-oriented proﬁle of a learner

and his/her desired knowledge/skills space and per-

forms a semantic search among annotated courses

and learning materials available in Learning Mate-

rials Space. Learning Materials Space will also be

connected to external tools and communication chan-

CSEDU 2017 - 9th International Conference on Computer Supported Education

222

nels, online courses, open-source learning materials.

The PSA also uses the PSP to re-plan initially created

study plan on the performance of a learner in the pro-

cess of his/her studies. With the help of Learning Pro-

cess Monitor, the PSA analyzes the performance of a

learner based on real time tests and feedback, recog-

nizes corresponding gaps in knowledge and requests

for missing learning materials from the Learning Ma-

terials Space Manager.

For the ILSS there will be good to classify the

each learning modules. The planning of the learn-

ing path will be on the content level, as well as on

the content parts level. For the fulﬁlling knowledge

gaps, we have to annotate the whole learning content

(courses, study modules, assignments, etc.). Then the

system will adapt the content according to learner’s

feedback.

The ILSS should take into account extracted not

only knowledge background of the learner, but also

his personal ability for learning new material. Ac-

cording to that, we have to minimize the time for

studying and maximize the quality of the learning

content. The system should improve the logic of

the learning concepts’ presentation, taking into ac-

count personal speciﬁcs of the learner via manipula-

tion with the complexity level of the system. Also,

system should analyze learning progress and see what

is missing and try to ﬁll missing knowledge by search-

ing the connection between learning modules in the

course. This connection search is needed because it

will be easy to see on what step of the learning pro-

cess learner starts to make mistakes and what parts of

modules inﬂuenced that.

4 DISCUSSION

The best way to produce experts is to accentuate the

learner’s best abilities and skills, assess the learner’s

potential and develop it further. That is why we badly

need revolutionary methods to facilitate intelligent

personalization of study processes and approaches

to make innovative education content more attrac-

tive and motivational for the learner (Khriyenko and

Khriyenko, 2013).

To be able to provide appropriate recommenda-

tion or suggestion of relevant content, the system

should be aware of learner preferences, experience

and knowledge, and goals. Therefore, availability of

recently updated learner proﬁle is one of the crucial

issues. The process of personal learner proﬁle cre-

ation (extension, modiﬁcation) should not be annoy-

ing for the learner, should not be time-consuming and

should not demand signiﬁcant efforts for him/her. The

main questions remain the same: how to distinguish

relevant parts of a proﬁle to be modiﬁed, how to au-

tomate and simplify the process of proﬁle manage-

ment as an integral part of other processes/activities

performed by the learner. Therefore, we are go-

ing to elaborate content-driven approach for personal

learner proﬁle management.

Having semantically annotated content and per-

sonalized knowledge/information spaces deﬁned by

correspondent semantic proﬁles of the learners, we

may populate the spaces with relevant content. Se-

mantic Web annotations enhance information re-

trieval and improve interoperability. Unfortunately,

human annotators are prone to error, and non-trivial

annotations usually require experts in the speciﬁc do-

main. Moreover, that is why annotations should be

maintained. If annotations are done cost-effectively,

then the future for the technology will not be limited.

The main idea behind proposed system is taking

into account knowledge and skills of the person. That

is why the whole learning content, courses, and learn-

ing materials should be annotated. For the annota-

tion, we chose semantic annotation in the form of on-

tology. The majority of the universities have to fol-

low the proposed ontology for it is better working

and using. In consequence, we can have a full set

of courses from different universities, available bach-

elor’s and master’s programs with speciﬁc require-

ments (i.e. with certain courses and backgrounds).

Learner, using such system will be able to include his

full transcript of records, then to ﬁnd a university, link

to courses that are interesting to him, as well as to ﬁnd

the shortest path for getting the higher education by

the present knowledge background.

The proposed ILSS makes possible to review and

reﬂect long-term information. One can produce a dif-

ferent kind of analyses related to individual strengths

and weaknesses. With the system, one can get ad-

ditional resources and individual recommendations

about learning technologies, materials, courses and

support.

A learner can use the system as a virtual career

agent. One can get information about available ca-

reers and education possibilities, and review personal

information against the prerequisites. One can eas-

ily produce an application, resume or curriculum vi-

tae and decide what information to include. This will

make the process of work applying more straightfor-

ward. The objective is that a learner can get com-

plete information related to own educational history

and up-to-date information related to work and career

possibilities.

For a teacher, the system will provide informa-

tion about the learner group. The teacher can get

An Intelligent Learning Support System

223

up-to-date and individual information about the per-

son’s learning, no matter where the learner is coming.

This will help the teacher to guide learners and per-

sonalize teaching for individual needs. The teacher

can produce different kind of analysis about the learn-

ers and get information about what methods, mate-

rials, and learning environments could be beneﬁcial

for them. The objective is that a teacher can get rel-

evant and up-to-date information about learners, cur-

rent research knowledge from the system, help with

difﬁculties, support practices and methods related to

learning.

This plan tries to fulﬁll the needs of modern digi-

tal and global society where learning and teaching are

not necessarily bounded to single institutions. The

concept of individual know-how will extend from de-

grees to a complete picture where formal education

is only one part. Further study and experiments will

consist of an extension of ontology for the University

of Jyv¨askyl¨a. This University has an in-house devel-

oped, integrated study information system

Korppi

which already has some annotation about courses, and

study modules. Because of that, it would be easy

to transfer presented information in the well struc-

tured knowledge base. As well as at the University of

Jyv¨askyl¨a is possible to transfer courses and credits

from other universities or online learning platforms.

5 CONCLUSIONS

Usually, computer based training or computer aided

learning does not use the model of the learner’s

knowledge and does not take into account personal-

ity of the learner, but it just uses traditional instruc-

tional methods (Phobun and Vicheanpanya, 2010).

The combination of adaptation, intelligence and per-

sonalization could improve the learning process.

In Web-based learning environments, it is an ad-

vantage to monitor the learner’s knowledge level and

automatically adjusts the content for each learner to

improve the education process and make learning in-

dividual for each learner or group of learners. Since

the Internet offers a vast amount of information, it is

important to help the teacher in the creation of the ma-

terials most suitable for education; and as well help

in ﬁnding the most relevant content and convert it

to comprehensive information. Also, a big privilege

of these Web-based learning environments is to help

learners with selecting courses and study programs

basing on their objectives, which will be beneﬁcial

in their further career development.

https://www.jyu.ﬁ/itp/en/korppi-guide

This paper describes the issues of the lifelong

learning, and how to make it more personalized,

supported by technology. Paper presents Intelligent

Learning Support System and describes it is main

modules.

An ontology to serve as the knowledge base of

ILSS was developed. All semantically annotated

information will be stored there for future SWRL-

formulated rules which allow extraction of new facts

from existing knowledge. The developed ontology is

very ﬂexible, and it is possible to modify data easily.

ILSS is connected to the different open-source

learning materials, external tools and communication

channels, and online courses. Also, external systems

and services with personalized user proﬁles can pro-

vide better guidance for a learner in a personalized

manner, as well as suggest new relevant content or

events which could help to ﬁll learner’s knowledge

gaps.

The proposed system might be very useful for uni-

versities and companies. It could also help in monitor-

ing changes in labor market and modiﬁcation of study

paths according to the called-for competence, as well

as to help in career planning.

REFERENCES

Aroyo, L. and Dicheva, D. (2004). The new challenges

for e-learning: The educational semantic web. Edu-

cational Technology & Society, 7(4):59–69.

Bendakir, N. and A¨ımeur, E. (2006). Using association

rules for course recommendation. In Proceedings of

the AAAI Workshop on Educational Data Mining, vol-

ume 3.

Berners-Lee, T. (2006). Linked data. design issues for

the world wide web. World Wide Web Consortium.

http://www. w3. org/DesignIssues/LinkedData. html.

Berners-Lee, T., Hendler, J., Lassila, O., et al. (2001). The

semantic web. Scientiﬁc american, 284(5):28–37.

Brusilovsky, P. and Peylo, C. (2003). Adaptive and in-

telligent web-based educational systems. Interna-

tional Journal of Artiﬁcial Intelligence in Education

(IJAIED), 13:159–172.

Canales, A., Pe˜na, A., Peredo, R., Sossa, H., and Guti´errez,

A. (2007). Adaptive and intelligent web based ed-

ucation system: Towards an integral architecture

and framework. Expert Systems with Applications,

33(4):1076–1089.

Chen, C.-M. (2008). Intelligent web-based learning system

with personalized learning path guidance. Computers

& Education, 51(2):787–814.

Dicheva, D., Mizoguchi, R., and Greer, J. E. (2009). Se-

mantic web technologies for e-learning, volume 4. Ios

Press.

Dolenc, K. and Aberˇsek, B. (2015). Tech8 intelligent

and adaptive e-learning system: Integration into tech-

CSEDU 2017 - 9th International Conference on Computer Supported Education

224

nology and science classrooms in lower secondary

schools. Computers & Education, 82:354–365.

Garc´ıa-Barrios, V. M., M¨odritscher, F., and G¨utl, C. (2005).

Personalisation versus adaptation? a user-centred

model approach and its application. In Proceedings

of the International Conference on Knowledge Man-

agement (I-KNOW), pages 120–127.

Gavriushenko, M., Kankaanranta, M., and Neittaanm¨aki, P.

(2015). Semantically enhanced decision support for

learning management systems. In Semantic Comput-

ing (ICSC), 2015 IEEE International Conference on,

pages 298–305. IEEE.

Gavriushenko, M., Khriyenko, O., and Porokuokka, I.

(2016). Adaptive vocabulary learning environment

for late talkers. In CSEDU 2016: Proceedings of

the 8th International Conference on Computer Sup-

ported Education. Vol. 2, ISBN 978-989-758-179-3.

SCITEPRESS.

Heath, T. and Bizer, C. (2011). Linked data: Evolving the

web into a global data space. Synthesis lectures on the

semantic web: theory and technology, 1(1):1–136.

Henderson, L. K. and Goodridge, W. (2015). Adviseme:

An intelligent web-based application for academic ad-

vising. International Journal of Advanced Computer

Science & Applications, 1(6):233–243.

Huang, M.-J., Huang, H.-S., and Chen, M.-Y. (2007). Con-

structing a personalized e-learning system based on

genetic algorithm and case-based reasoning approach.

Expert Systems with Applications, 33(3):551–564.

Jeong, H.-Y., Choi, C.-R., and Song, Y.-J. (2012). Person-

alized learning course planner with e-learning dss us-

ing user proﬁle. Expert Systems with Applications,

39(3):2567–2577.

Jovanovic, J., Gasevic, D., Brooks, C., Devedzic, V., Hatala,

M., Eap, T., and Richards, G. (2007). Using semantic

web technologies to analyze learning content. IEEE

Internet Computing, 11(5).

Jovanovi´c, J., Gaˇsevi´c, D., and Devedˇzi´c, V. (2009). Tan-

gram for personalized learning using the semantic

web technologies. Journal of emerging technologies

in web intelligence, 1(1):6–21.

Jovanovi´c, J., Gaˇsevi´c, D., Knight, C., and Richards, G.

(2007). Ontologies for effective use of context in e-

learning settings. Educational Technology & Society,

10(3):47–59.

Jyothi, N., Bhan, K., Mothukuri, U., Jain, S., and Jain, D.

(2012). A recommender system assisting instructor in

building learning path for personalized learning sys-

tem. In Technology for Education (T4E), 2012 IEEE

Fourth International Conference On, pages 228–230.

IEEE.

Khriyenko, O. and Khriyenko, T. (2013). Innovative edu-

cation environment and open data initiative: Steps to-

wards user-powered society-oriented systems. GSTF

Journal on Computing (JoC), 3(3):31.

Leung, C. M., Tsang, E. Y., Lam, S., and Pang, D. C.

(2010). Intelligent counseling system: A 24 x 7 aca-

demic advisor. Educause Quarterly, 33(4):n4.

Myneni, L. S., Narayanan, N. H., Rebello, S., Rouinfar, A.,

and Pumtambekar, S. (2013). An interactive and in-

telligent learning system for physics education. IEEE

Transactions on learning technologies, 6(3):228–239.

Nguyen, T. B., Nguyen, D. N., Nguyen, H. S., Tran, H., and

Hoang, T. A. D. (2008). An integrated approach for

an academic advising system in adaptive credit-based

learning environment.

Nurjanah, D. (2016). Good and similar learners’ recom-

mendation n adaptive learning systems. pages 434–

440.

Phobun, P. and Vicheanpanya, J. (2010). Adaptive in-

telligent tutoring systems for e-learning systems.

Procedia-Social and Behavioral Sciences, 2(2):4064–

4069.

Ranganathan, G. R. and Brown, J. A. (2007). The use

of ontologies and rules to assist in academic advis-

ing. In International Workshop on Rules and Rule

Markup Languages for the Semantic Web, pages 207–

214. Springer.

Shvaiko, P. and Euzenat, J. (2013). Ontology matching:

state of the art and future challenges. IEEE Transac-

tions on knowledge and data engineering, 25(1):158–

176.

Uren, V., Cimiano, P., Iria, J., Handschuh, S., Vargas-Vera,

M., Motta, E., and Ciravegna, F. (2006). Semantic an-

notation for knowledge management: Requirements

and a survey of the state of the art. Web Semantics:

science, services and agents on the World Wide Web,

4(1):14–28.

Wen, F. L. S. L. D. and McGreal, F. Z. K. R. (2007). e-

advisor: A multi-agent system for academic advising.

Werghi, N. and Kamoun, F. K. (2009). A decision-

tree-based system for student academic advising and

planning in information systems programmes. In-

ternational Journal of Business Information Systems,

5(1):1–18.

An Intelligent Learning Support System

225