ICALTS JEIRP Interaction

1
ICALTS JEIRP Interaction Collaboration
Analysis Supporting Teachers Students
Self-regulation
www.rhodes.aegean.gr/LTEE/kaleidoscope-icalts
2
JEIRP ICALTS ? Identity / Partners
? Learning Technology and Educational Engineering
LTEE Laboratory, University of the Aegean,
Greece, Angelique Dimitracopoulou ?
Collaborative and Learning Support Systems,
COLLIDE, University of Duisburg, German, Ulrich
Hoppe ? Center for Research Support for
Training and its Technologies CRAFT, /EPFL,
Switzerland, Pierre Dillenbourg ? Learning
Technologies and Collaborative Systems Group,
UNED, Spain, Felisa Verdejo ? EDTE, University
of Twente, Netherlands, Ton de Jong ? Cognitive
Science and Educational Technology Laboratory
(COSET), University of Athens, Greece, Stella
Vosniadou ? InterMedia, University of Oslo,
Norway, Anders Morch ? ICAR Group, CNRS, France,
Michael Baker

• A complementary group
• Cognitive psychologists,
• Engineers,
• Designers in specific domains
• with expertise in a variety of systems

3
JEIRP ICALTS ? Identity / resources

• Duration 10 months year 2004
• ? Budget 25.000 E

Budget for travel expenses Three meetings
? Grenoble, March 04 ? Rhodes, July,
04 ? Lausanne, October 04
4
JEIRP ICALTS ? Purpose
ICALTS purpose is to put the accent on the
notion of the support that participants (students
but also teachers) need during individual or
collaborative learning activities. Such as
support could be provided via the use of
appropriate Interaction Analysis IA tools that
analyse participants interactions.
Interaction Analysis Tools could support ?
student, groups awareness, metacognition,
self-assessment ?selfregulation of their
own activity ? teacher(s) diagnosis,
assessment ? teachers strategies adaptation
during Technology-Based Learning Activities.
5
JEIRP ICALTS ? Vision
A new research subfield
Interactions Analysis
Intelligent Systems Guiding Students
Researchers
Computer-based analysis offering

• A new research subfield
• Computer based Interaction Analysis
• filtering and processing of the interactions
  and the product of
• the activity

6
JEIRP IA ? Necessity-Interest

• Designers researchers havent taken into
  account
• ? The difficulties of teachers when they have to
  manage activities in collaborative environments
  or in stand-alone systems, due to the very
  complexe interactions that occur.
• Students cannot create an image of their own
  activity, or this of other students/collaborators
  (as individuals, group or community).

? In order to integrate technology based learning
environments in Education Support the interest
to use technology based learning environments, in
every day educational activities, where there is
a need to (self)evaluate in an operational way
both the learning outcomes/processes and the
quality of collaboration.
7
JEIRP ICALTS ? IA tools examples
Awareness functions addressed to the
administrator or the teacher
phpBB Forum statistics Number of
posts Number of users Database size Webct,
Summary of activity report Most active day,
Most active hour per day, Average users per
day, etc.
Awareness functions addressed to students
8
JEIRP ICALTS ? IA toolsexamples
9
JEIRP ICALTS ? IA tool generic

• ? To what the data are referred?
• the actions or the product of
• individuals students,
• -a specific group of students,
• -a number of students group,
• -a teacher, etc
• (b) the actions of a teacher

• ? To whom are addressed to (which is the user)?
• the student,
• a group of students,
• the teacher,
• the system it-self,
• the researcher

• ? Interaction analysis tools
• awareness tools
• assessment tools
• guiding tools-advisors
• mixed tools

norme
10
JEIRP ICALTS ? A central Concept

• ? A central concept Interaction Analysis
  Indicator
• (variable which represent one aspect of the
  interaction)
•  awareness
• (e.g. answered contributions per agent,
  learners ratio participation,
• degree of presence in the forum,
  vicarious interactionread-writing),
•  mode or quality of the contribution
• (p.ex. activity level indicator per
  category of posting,
• work amount, argumentation level, )
•  mode or quality of collaboration 
• (p.ex. interaction level, initiative,
  contribution to group process,
• actors degree centrality, network
  density),

11
(No Transcript)
12
JEIRP ICALTS ? Current state/indicators
Indicator Interaction Model
In an ideal case, for each learning session, a
number of complementary indicators could be
produced. Differents indicators combinations
could forme a descriptive model or even
evaluation model according to specific
categories of learning activities. For this
purpose, it is necessary to attribute a
psychological/pedagogiqual meaning to most of the
indicators.
gt Current state of available indicators -
the defined indicators do not forme coherent
set(s) - most of them are low level
indicators
13
JEIRP ICALTS ?Project Results/D1State of the
art
?A Need to focus on indicators produced by the
consortium members Challenge Find a common
descriptive language, making appear hidden
aspects

• ? ICALTS/ D.26.1.  State of the Art on
  Interaction Analysis
• Interaction
  Analysis Indicators 
• ? A Framework on IA Indicator Concept
• ? A list of described applied indicators
  gt 50 indicators

14
JEIRP ICALTS ? Project Results /D2

• Unified framework
• based on a general schema of interaction
  analysis
• ? Classification description of learning
  environments, regarding
• The user cognitive system individual, duad,
  small group, group, community
• The mode of interaction face to face, computer
  mediated
• The temporal organisation of interaction syn,
  asyn
• Task types simulation, modelling,
  argumentation, text production, etc),
• ? Classification /description of Analysis methods
• ? Classification of IA tools

Integration of different aspects ? Learning
environments and contexts of use stand
alone use, collaborative use,
synchronous collaborative, asychronous one,
for a restricted or a wide community ?
Participants and participants profile
students, teachers their various profiles
roles
? ICALTS/ D.26.2.  Unified Framework of
Interactions Analysis
15
JEIRP ICALTS ?D3 Research Directions
Indicators

• Focus, work, in a profound way on IA
  indicators
• There are differences or even gaps among
  
• - the indicators that we process
• - these having a psychological/pedagog
  ical significance
• - those that can activate a
  metacognitive process to users

psycho/pedagogical Indicators
-researchers-
Processed Indicators
meta-cognitive indicators
-users-
16
JEIRP ICALTS ? D3 Research Directions

• Knowing better the IA User
• Different Indicators and/or different
  visualisations
• taking into account ? the
  profiles and
• 
  the roles of users
• 
  ? the type of the activity
• 
  ? the context
• Workshop Lausanne, 2004  Tailoring IA
  indicators for different types of users 

Indicators
17
JEIRP ICALTS ? D3 Research Directions
IA methods
? Explore-apply AI techniques Advanced data
mining methods

• Artificial Intelligence (AI) techniques such as
• ?pattern ? plan recognition and/or ? data
  mining
• may be valuable in the construction of
  indicators from raw interaction data.
• For instance
• Plan recognition techniques to identify
  different problem-solving strategies,
• Markov modeling to recognize or predict
  interaction patterns,
• Pattern recognition techniques to allow
  sequences of events to be grouped
• into more general behavioral units,
• Bayesian modeling to describe the relationships
  between actions causal effects
• Filtering techniques to determine which actions
  are meaningful, and
• which should be disregarded as noise.

gt New JEIRP  Design Patterns  possible
contribution
18
JEIRP ICALTS ? D3 Research Directions
IA methods
? Develop a unified approach of IA methods
The techniques and systems use different
standards for representing raw data, and
defining analysis methods. ? There is a need to
propose shared and agreed upon data formats and
interfaces for analysis methods. ? There is a
need to develop modular, reusable applications
allowing researchers to share and reuse tools in
different learning environments.
gt New JEIRP  Interactions Analysis  possible
contribution
19
JEIRP ICALTS ? .. and Kal NoE Integration ?

• Other JPAs to take into account or to be taken
  into account by
• SIGs CSCL, AIED, CSIL Dissemination
• ERTs TELMA, ERT on Collaborative Scripts
• JEIRPs MOSIL, Trails, // new Design Patterns

Backbone activities to contribute ? ATA
(advanced training activities) ?  VDS (Virtual
Doctoral School) ?  VL (Virtual Laboratory)
20
JEIRP ICALTS
www.rhodes.aegean.gr/LTEE/kaleidoscope-icalts