Skaidre 1

1
Application of Intelligent Technologies in
Computer Engineering Education
Assoc. Prof. Dr. Eugenijus Kurilovas Vilnius
University Institute of Mathematics and
Informatics
IFIP TC3 Conference. Vilnius. 2 July, 2015
2
Introduction
What learning content, methods and technologies
are the most suitable to achieve better learning
quality and efficiency? In Lithuania, we believe
that there is no correct answer to this question
if we dont apply personalised learning approach.
We strongly believe that one size fits all
approach doesnt longer work in education. It
means that, first of all, before starting any
learning activities, we should identify students
personal needs their preferred learning styles,
knowledge, interests, goals etc. After that,
teachers should help students to find their
suitable (optimal) learning paths learning
methods, activities, content, tools, mobile
applications etc. according to their needs.
But, in real schools practice, we cant assign
personal teacher for each student. This should be
done by intelligent technologies. Therefore, we
believe that future school means personalisation
plus intelligence. In this presentation,
Lithuanian Intelligent Future School (IFS)
project is presented aimed at implementing both
learning personalisation and educational
intelligence.
3
Outline

• Related EU-funded Future Classroom Lab projects
• IFS concept and implementation vision research
  and development, application and validation of
  intelligent technologies in education
• IFS related RD works already done
• Conclusion

4
Related Future Classroom Lab projects
5
http//itec.eun.org/

• iTEC (Innovative Technologies for Engaging
  Classrooms) 2010-2014, 7FP
• How did the iTEC approach impact on learners and
  learning
• Key finding 1 Teachers perceived that the iTEC
  approach developed students 21st century skills,
  notably independent learning critical thinking,
  real world problem solving and reflection
  communication and collaboration creativity and
  digital literacy. Their students had similar
  views.
• Key finding 2 Student roles in the classroom
  changed they became peer assessors and tutors,
  teacher trainers, co-designers of their learning
  and designers/producers.
• Key finding 3 Participation in classroom
  activities underpinned by the iTEC approach
  impacted positively on students motivation.
• Key finding 4 The iTEC approach improved
  students levels of attainment, as perceived by
  both teachers (on the basis of their assessment
  data) and students.

6



http//lsl.eun.org/
LSL (Living Schools Lab) 2012-2014, 7FP With
the participation of 15 partners, including 12
education ministries, LSL project promoted a
whole-school approach to ICT use, scaling up best
practices in the use of ICT between schools with
various levels of technological proficiency. The
participating schools were supported through
peer-exchanges in regional hubs, pan-European
teams working collaboratively on a number themes,
and a variety of opportunities for teachers'
ongoing professional development. Observation of
advanced schools in 12 countries produced a
report and recommendations on the mainstreaming
of best practice, and the development of
whole-school approaches to ICT.
7

http//creative.eun.org/

• CCL (Creative Classrooms Lab, CCL) 2013-2015,
  LLP
• CCL brought together teachers and policy-makers
  in 8 countries to design, implement and evaluate
  11 tablet scenarios in 45 schools. CCL produced
  learning scenarios and activities, guidelines and
  recommendations to help policy-makers and schools
  to take informed decisions on optimal strategies
  for implementing 11 initiatives in schools and
  for the effective integration of tablets into
  teaching and learning.
• The 11 computing paradigm is rapidly changing,
  particularly given the speed with which tablets
  from different vendors are entering the consumer
  market and beginning to impact on the classroom.
  Over the next 2-3 years policy makers will face
  some difficult choices How to invest most
  efficiently in national 11 computing programmes?
  What advice to give to schools that are
  integrating tablets?
• To address these challenges, CCL carried out a
  series of policy experimentations to collect
  evidence on the implementation, impact and
  up-scaling of 11 pedagogical approaches using
  tablets. Lessons drawn from the policy
  experimentations also
• Provide guidelines, examples of good practice and
  a training course for schools wishing to include
  tablets as part of their ICT strategy.
• Support capacity building within Ministries of
  Education and regional educational authorities
  and encourage them to introduce changes in their
  education systems.
• Enable policy makers to foster large-scale uptake
  of the innovative practice that is observed
  during the project.

8
IFS concept
9
5 Empower Redefinition innovative use Technology supports new learning services that go beyond institutional boundaries. Mobile and locative ICT support agile teaching and learning. The learner as a co-designer of the learning journey, supported by intelligent content and analytics.
4 Extend Network redesign embedding Ubiquitous, integrated, seamlessly connected ICT support learner choice and personalisation beyond the classroom. Teaching and learning are distributed, connected and organised around the learner. Learners take control of learning using ICT to manage their own learning
3 Enhance Process redesign Teaching and learning redesigned to incorporate ICT, building on research in learning and cognition. Institutionally embedded ICT supports the flow of content and data, providing an integrated approach to teaching, learning and assessment. The learner as a producer using networked ICT to model and make.
2 Enrich Internal Coordination ICT used interactively to make differentiated provision within the classroom. ICT supports a variety of routes to learning. The learner as a user of ICT tools and resources
1 Exchange Localised use ICT is used within current teaching approaches. Learning is teacher-directed and classroom-located. The learner as a consumer of learning content and resources
___________________________________________
10

• Future school means personalisation plus
  intelligence
• IFS implementation stages
• (based on iTEC schools innovation maturity
  model)
• Creating learners models (profiles) based on
  their learning styles and other particular needs
• Interconnecting learners models with relevant
  learning components (learning content, methods,
  activities, tools, apps etc.) and creating
  corresponding ontologies
• Creating intelligent agents and recommender
  systems
• Creating and implementing personalised learning
  scenarios (e.g. in STEM Science, Technology,
  Engineering and Mathematics subjects)
• Creating educational multiple criteria decision
  making models and methods

11
Personalisation
12
Personalisation creating students profiles

• Selecting good taxonomies (models) of learning
  styles, e.g., (Felder Silverman, 1988), (Honey
  Mumford, 2000), the VARK style (Fleming, 1995)
• Creating integrated learning style model which
  integrates characteristics from several models.
  Dedicated psychological questionnaire(s)
• Creating open learning style model
• Using implicit (dynamic) learning style modelling
  method
• (5) Integrating the rest features in the student
  profile (knowledge, interests, goals)

13
Personalisation identifying learning styles
14
Personalisation identifying learning styles

• VARK inventory was designed by Fleming in 1987
  and is an acronym made from Visual, Aural,
  Read/write and Kinaesthetic. These modalities are
  used for preferable ways of learning (taking and
  giving out) information
• Visual learners prefer to receive information
  from depictions in figures in charts, graphs,
  maps, diagrams, flow charts, circles,
  hierarchies, and others. It does not include
  pictures, movies and animated websites that
  belong to Kinaesthetic.
• The aural perceptual mode describes a preference
  for spoken or heart information. Aural learners
  learn best by discussing, oral feedback, email,
  chat, discussion boards, and oral presentations.
• Read/write learners prefer information displayed
  as words quotes, lists, texts, books, and
  manuals.
• The kinaesthetic perceptual mode describes a
  preference for reality and concrete situations.
  They prefer videos, teaching others, pictures of
  real things, examples of principles, practical
  sessions, and others.
• Multimodals are those learners who have
  preferences in more than one mode.

15
Creating recommender system
Learning styles (Honey and Mumford, 1992) Preferred learning activities Suitable teaching / learning methods (iCOPER D3.1, 2009)) Suitable LO types (LRE AP v4.7, 2011)
Activists are those people who learn by doing. Have an open-minded approach to learning, involving themselves fully and without bias in new experiences Brainstorming, problem solving, group discussion, puzzles, competitions, and role-play Active Learning, Blogging, Brainstorming and Reflection, Competitive Simulation, E-Portfolio, Creation of Personalised Learning Environments, Creative Workshops, Exercise Unit, Games Genre, Presenting Homework, Image Sharing, In-class Online Discussion, Mini Conference, Modelling, Online Reaction Sheets, Online Training, Peer Assessment, Process-based Assessment, Process Documentation, Project-based Learning, Resource-based Analysis, Role Play, Student Wiki Collaboration, World Café, Web Quest Application, Assessment, Broadcast, Case study, Drill and practice, Educational game, Enquiry-oriented activity, Experiment, Exploration, Glossary, Open activity, Presentation, Project, Reference, Role play, Simulation, Tool, Website
16
Creating recommender system
17
Creating recommender system
18
Creating recommender system
iOS (Apple iPad) Android (Samsung) iOS / Android Suitable LO types
Idea Sketch lets you easily draw a diagram mind map, concept map, or flow chart - and convert it to a text outline, and vice versa. You can use Idea Sketch for anything, such as brainstorming new ideas, illustrating concepts, making lists and outlines, planning presentations, creating organizational charts, and more Mindjet for Android rated as one of the best mind mapping apps for Android. Create nodes and notes, add images of your own or icons provided, and add attachments and hyperlinks. Sync to your Dropbox Mind Mapping lets you create, view and edit mind maps online or offline and lets the app synch with your online account whenever connected. You can share mind maps directly from the device, inviting users via email. You can add icons, colours and styles, view notes, links and tasks and apply map themes, drag and drop and zoom Application, Broadcast, Enquiry-oriented activity, Glossary, Open activity, Presentation, Reference, Role play, Simulation, Tool, Website
Interconnection of Activists Brainstorming
learning activity with suitable apps and LOs
types
19
Creating recommender system
20
Example Integrating Web 2.0 tools into learning
activities
21
Recommender systems (as a kind of services in the
e-learning environment) can provide personalised
learning recommendations to learners.
Recommender systems are information processing
systems that gather various kinds of data in
order to create their recommendations. The data
are primarily about the items (objects that are
recommended) to be suggested and the users who
will receive these recommendations. The data
can be formalised in domain ontology, thus the
knowledge about a user and items becomes reusable
for people and software agents. Also, the
ontology could contain a useful knowledge that
can be used to infer more interests than can be
seen by just an observation. The aim of TEL is
to improve learning. It is therefore an
application domain that generally covers
technologies that support all forms of learning
activities. An important activity in TEL is
search-ability relevant learning resources and
services as well as their better finding.
Recommender systems support such an information
retrieval.
22
There are different types of recommender systems
based on the recommendation approaches
content-based, collaborative filtering,
demographic, knowledge-based, community-based,
utility-based, hybrid, and semantic. In this
research, knowledge-based recommender system
using rules-based reasoning is used.
Knowledge-based systems recommend items based on
the specific domain knowledge about how certain
item features satisfy users needs and
preferences as well as how the item is useful for
the user. Knowledge-based recommender systems
can be rule-based or case-based. The form of data
collected by the knowledge-based system about
users preferences can be statements, rules, or
ontologies. The knowledge base of the
rule-based system comprises the knowledge that is
specific to the domain of the application. The
rule-based reasoning system represents knowledge
of the system in terms of a bunch of rules
(facts). These rules are in the form of IF THEN
rules such as IF some condition THEN some
action. If the condition is satisfied, the
rule will take the action.
23

• The proposed method for Web 2.0 tools integration
  into learning activities is based on the ontology
  developed.
• With the view to find a particular Web 2.0 tool
  suitable for the accomplishment of the learning
  activity, a link between the tool and the
  learning activity must be identified. This
  relationship can be established by
  interconnections between the defined tool and
  activity elements.
• The learning activity is defined as consisting of
  the following elements
• Learning Activity (what action a learner
  performs)
• Content (which object a learner manages)
• Interaction (with whom a learner interacts) and
  
• Synchronicity (at what time a learner performs
  the intended action).
• Web 2.0 tool is defined as set of universal
  functions. This universal function is defined as
  consisting of the following elements
• Function (what action can be performed by using
  a tool)
• Artefact (which object can be managed by using a
  tool)
• Interaction (what kind of interaction the tool
  enables) and
• Synchronicity (at what time the intended action
  is enabled by a tool to take place).

24
The Learning activities and Functions of tools
are classified mostly based on the Conole, 05
media taxonomy. These types and particular
elements are presented in Table 2
Type Learning activities Subtype (1-8) Web 2.0 tool function
Narrative Revise 1 View Explore ( Read, view, listen)
Information management Find 2 Search Search
Information management Collect 3 Host Host (Store), Syndicate
Productive Prepare 4 Create Create (draw, write, record, edit)
Communicative Present 5 Share Share, publicise
Communicative Dispute 6 Discuss Communicate
Imitative Role play 7 Imitate Simulate (Game simulation)
Imitative Observation 8 Model Model (Phenomenon modelling)
Table 2 Learning activities and Web 2.0 tools
functions types
25
Thus, Web 2.0 tools could be divided based on
their usage possibilities, managed objects,
communication form, and sort of imitation process
into three groups as follows (1) Artefacts
management, (2) Communication, and (3) Imitation
tools. We have defined the following components
in the domain ontology visualised with Protégé
4.3 ontology editor Concepts (Main Classes)
(Figure 1), and Relationships between Concepts
(Properties) (Figure 2)
26

• The stages of the method of integrating Web 2.0
  tools into learning activities are as follows
• Identification of learners learning style (i.e.
  preferences of the learning content and
  communication modes)
• Selection of the learning objective and the
  learning method
• Determination of the elements of chosen learning
  method activities
• Determination of universal function elements of
  each Web 2.0 tool
• Finding of the link between tool and learning
  activity elements
• Selection of a suitable tool based on specified
  elements Action, Interaction, Synchronicity.
  Artefact is determined based on individual
  learning style.
• Description of each stage and the detailed
  presentation of the method are provided in
  Juskeviciene, Kurilovas, 14.

27
In order to ascertain the suitability of this
approach, the recommender system prototype was
developed. This prototype was developed following
the working principles of the knowledge-based
recommender system. The domain knowledge was
conceptualised in the ontology. The prototype
of the knowledge-based recommender system
implements this method completely
Scheme of the recommender system
28
Recommender system prototype operation
29
Example educational multiple criteria decision
making
30
Multiple Criteria Decision Making Scalarisation
method the experts additive utility function
The major is the meaning of the utility function
the better LOs meet the quality requirements in
comparison with the ideal (100) quality
According to scalarisation method, we need LOs
evaluation criteria ratings (values) and weights
31
Linguistic variables conversion into triangle
non-fuzzy values and weights Linguistic
variables Non-fuzzy values Excellent /
Extremely valuable 0.850 Good / Very
valuable 0.675 Fair / Valuable 0.500 Poor /
Marginally valuable 0.325 Bad / Not
valuable 0.150
32

• In identifying quality criteria for the decision
  making, the following considerations are relevant
  to all multiple criteria decision making
  approaches
• Value relevance
• Understandability
• Measurability
• Non-redundancy
• Judgmental independence
• Balancing completeness and conciseness
• Operationality
• Simplicity versus complexity

33

34
Papers 2015

• Kurilovas, E. Juskeviciene, A. Bireniene, V.
  (2015). Research on Mobile Learning Activities
  Using Tablets. In Proceedings of the 11th
  International Conference on Mobile Learning (ML
  2015). Madeira, Portugal, March 1416, 2015, pp.
  9498.
• Kurilovas, E. Zilinskiene, I. Dagiene, V.
  (2015). Recommending Suitable Learning Paths
  According to Learners Preferences Experimental
  Research Results. Computers in Human Behavior
  in print, doi10.1016/j.chb.2014.10.027 Q1
• Kurilovas, E. Juskeviciene, A. (2015). Creation
  of Web 2.0 Tools Ontology to Improve Learning.
  Computers in Human Behavior in print,
  doi10.1016/j.chb.2014.10.026 Q1
• Kurilovas, E. Vinogradova, I. Kubilinskiene, S.
  (2015). New MCEQLS Fuzzy AHP Methodology for
  Evaluating Learning Repositories A Tool for
  Technological Development of Economy.
  Technological and Economic Development of Economy
  in print Q1
• Kurilovas, E. (2015). Future School
  Personalisation plus Intelligence. Chapter in
  Handbook of Research on Information Technology
  Integration for Socio-Economic Development. IGI
  Global in print

35
Papers 2014

• Kurilovas, E. Juskeviciene, A. Kubilinskiene,
  S. Serikoviene, S. (2014). Several Semantic Web
  Approaches to Improving the Adaptation Quality of
  Virtual Learning Environments. Journal of
  Universal Computer Science, Vol. 20 (10), 2014,
  pp. 14181432.
• Kurilovas, E. Kubilinskiene, S. Dagiene, V.
  (2014). Web 3.0 Based Personalisation of
  Learning Objects in Virtual Learning
  Environments. Computers in Human Behavior, Vol.
  30, 2014, pp. 654662. Q1
• Kurilovas, E. Zilinskiene, I. Dagiene, V.
  (2014). Recommending Suitable Learning Scenarios
  According to Learners Preferences An Improved
  Swarm Based Approach. Computers in Human
  Behavior, Vol. 30, 2014, pp. 550557. Q1
• Kurilovas, E. Serikoviene, S. Vuorikari, R.
  (2014). Expert Centred vs Learner Centred
  Approach for Evaluating Quality and Reusability
  of Learning Objects. Computers in Human Behavior,
  Vol. 30, 2014, pp. 526534. Q1
• Juskeviciene, A. Kurilovas, E. (2014). On
  Recommending Web 2.0 Tools to Personalise
  Learning. Informatics in Education, Vol. 13 (1),
  2014, pp. 1730
• Kurilovas, E. (2014). Research on Tablets
  Applications for Mobile Learning Activities.
  Journal of Mobile Multimedia, Vol. 10 (34),
  2014, pp. 182193.

36
Papers 2013

• Kurilovas, E. Serikoviene, S. (2013). New MCEQLS
  TFN Method for Evaluating Quality and Reusability
  of Learning Objects. Technological and Economic
  Development of Economy, Vol. 19 (4), 2013, pp.
  706723. Q1
• Kurilovas, E. Zilinskiene, I. (2013). New MCEQLS
  AHP Method for Evaluating Quality of Learning
  Scenarios. Technological and Economic Development
  of Economy, Vol. 19 (1), 2013, pp. 7892. Q1
• Kurilovas, E. (2013). MCEQLS Approach in
  Multi-Criteria Evaluation of Quality of Learning
  Repositories. Chapter 6 in the book José Carlos
  Ramalho, Alberto Simões, and Ricardo Queirós
  (Ed.) Innovations in XML Applications and
  Metadata Management Advancing Technologies. IGI
  Publishing, USA, 2013, pp. 96117.
• Kurilovas, E. Serikoviene, S. (2013). On
  E-Textbooks Quality Model and Evaluation
  Methodology. International Journal of Knowledge
  Society Research, Vol. 4 (3), 2013, pp. 6678.

37
Papers 2012

• Kurilovas, E. Zilinskiene, I. (2012). Evaluation
  of Quality of Personalised Learning Scenarios An
  Improved MCEQLS AHP Method. International Journal
  of Engineering Education, Vol. 28 (6), 2012, pp.
  13091315.
• Kurilovas, E. Serikoviene, S. (2012). New TFN
  Based Method for Evaluating Quality and
  Reusability of Learning Objects. International
  Journal of Engineering Education, Vol. 28 (6),
  2012, pp. 12881293.
• Zilinskiene, I. Dagiene, V. Kurilovas, E.
  (2012). A Swarm-based Approach to Adaptive
  Learning Selection of a Dynamic Learning
  Scenario. In Proceedings of the 11th European
  Conference on e-Learning (ECEL 2012). Groningen,
  the Netherlands, October 2627, 2012, pp.
  583593.

38
IFS concept implementation vision
39

• Collaboration agreements between Vilnius
  University and (20 pilot) schools on IFS
  implementation
• Joint expert group on creating interconnections
  and intelligent agents
• RD, creation of technologies and scenarios, and
  validation at schools
• Feedback, questionnaires, interviews, data mining
• Return to (3) based on (4)

40
Conclusion
41

• Future school means personalisation
  intelligence
• Learning personalisation means creating and
  implementing personalised learning paths based on
  recommender systems and personal intelligent
  agents suitable for particular learners according
  to their personal needs
• Educational intelligence means application of
  intelligent technologies and methods enabling
  personalised learning to improve learning quality
  and efficiency
• Lithuanian IFS project is aimed at implementing
  both learning personalisation and educational
  intelligence

42

Welcome to collaborate. Thank you for your
attention. Questions? Dr. Eugenijus Kurilovas
http//eugenijuskurilovas.wix.com/my_site