Thursday 27 April 2017

Liveblog #Educon17 @Kinshuk1 Enhancing learning through adaptivity and personalization in ubiquitous environments

Kinshuk (http://www.kinshuk.info ) was streamed in live from Austin, Texas. Lately he is also increasingly engagement with industry on EdTech (yes, the bridge between university and industry is tightening).
The learning environment is expanding outside of the classroom environment, so how can we incorporate learning in all these environments. Some opportunities (free)
  • Series in springer collection in EdTech (look up book guidelines for this series: I think it is http://www.springer.com/series/11777 ), any new advancements are welcomed.
  •  Journal which is completely open access called ‘smart learning environments’ (Inge, look this up: http://www.springer.com/computer/journal/40561 ) , focus on improved learning environments, and bringing these traditional environments and transforming them into online learning environments.
  • International association of smart learning environments http://www.iaslo.net they look for evidence-based research on the subject.


Current trends in learning
·        Inclusive education,
·        Focus on individual strengths and needs
·        Various learning scenarios – in clsass and outdoor environments
·        Relevance of the learning scenarios with learners living and working environments
·        Authentic learning with physical as well as digital resources
Result: better learning experience due to authentic learning, and ubiquitous access to learning. So learning is now more easily fitted to real life of the learner. Learning needs to be relevant to the learner, but as a teacher you need to become aware of how to capture the attention outside of the classroom.
This means the teachers must become aware of the new teaching/learning opportunities.

Vision
Learning is happening everywhere, at any time, and is highly contextualised.
Seamless integration of learning into every aspect of life with implies immersive, always on learning that happens so naturally and in such small chunks that no conscious effort in needed by actively learning while engaged in education.
We need to make learning as meaningful as possible. The goal of the learning needs to be put across to all the learners, and the learning needs to be made visible (e.g. Hattie)… but all of this is highly demanding for the teacher. Every student is doing different things, so how can the teacher know that her learners are learning? That is why we are looking for much more data, much more information, and the assessments is also coming out of the classrooms and out of the formal, classic design of assignments and assessments.

Smart learning analytics is used to discover what type of learning data is coming in. Discover, analyze and make sense of student, instruction and environmental data from multiple sources to identify learning traces in order to facilitate instructional support in authentic learning environments. This also opens a new type of teaching, namely coaching, give guidance, personalise the feedback given the learner data or the learner information that is viewed and analysed by the teacher. For example, a  flower bed with a placard on what the flowers are, but on the top right there is a QR code with additional information on the flowers, but embedded in its full cycle, use and systematic botanical information. So this means that the information is delivered in an adaptive way (as complex as the learner wants to view it), and open to all. The learning system provides you authentic information within a contextual reality, and with the option to zoom in on additional information. (look at iSpot as additional learning scenario).
Information can now come from different sources: mobiles, environment, internet, people, …. It is like learning traces, a small learning impression that can tell us that learning is actually happening. For instance, looking at paintings in a museum, one painting captures the learners attention, and some things are different to other paintings. The learner might learn something a week later, and gets more information on it, and now a story can be shared by the learner to people that are outside of the classroom. This actual fact proofs that learning has happened.
But a system needs to be in place to proof or visualise the actual learning that is happening.

Remark on data: the learners need to be made aware that their data might be used, for privacy and policy issues.

How can we design instructional support that will make this type of smart learning happen and make it measurable.

Discover
Past record and real-time observation of: learner’s capabilities, preferences and competencies, learenr’s location, learner’s technological use, technologies surrounding the learner, changes keep happening in the learner’s situational context. So knowing the past, does not mean that what is happening today is a meaningful difference to the previous actions, as the contexts of today constantly change.
Miller was pioneering (5 elements of information memorisation).
And although the tech can provide the teachers with lots of additional data, the actual learning experience needs to take into account the changing environment and connected conditions of this environments.

Human-machine learning has an effect on the actual learning process.
Is the learner trying to find new information, is that new information screened critically…
We do have lots of mechanisms that we use to see what the learners are going through and how the learning occurs.
Informal learning happens everywhere, across the potential learning environments, and is there a record of the learning somewhere? Small learning can happen anywhere, but how can we identify it and use it as evidence of learning.

Making sense: learning traces
A learning trace comprises of a network of observed study activities that lead to a measurable cchunk of learning.
Learning traces are sensed ad supply data to learning analytics, where data is typically big, un/semi structured, seemingly unrelated, not quite truthful, and fits multiple models and theories.
What kind of learning, which models can be used to map learning traces to try to understand that learning is actually happening. Learning traces are also important to understand personalised learning, differentiated learning that is happening across the population in all its variety.

Why learning traces are important
Different students can adopt different learning approaches for the same learning activity
Ex,, why a pointed object penetrates better than a blunt object?
A visual-oriented learner may choose to use different approach than an sensoratory learner.

Learner awareness
Personalisation of learning experience through dynamic learner modeling: performance, meta-cognitive skills, cognitive skills, learning styles, affective state, physiological symptoms (eg. The learner is doing something in the lab, and suddenly heart rate will increase, why? What kind of concept is the learners using, are there comparable situations of learning where this occurred?). All of this are tools that can make teachers more informed, enabling more informed decisions on learning.

Technological awareness
Personalization of learning experience through the identification of technological functionality.
Identifying various device functionality
Dynamically optimize the content to suit the functionality
Display capability, audio and video capability…

Location awareness
Personalisation through location modelling
Location base optimal grouping (grouping ad hoc based on mobile location)
Location based adaptation of learning content

Real-life physical objects
Public databases of POIs
QR codes
Wifi and Bluetooth access point identification
Active and passive RFIDs

Surrounding awareness
Learning based on all the surrounding data, context-aware knowledge structures
Identifying specific context-aware knowledge structure among different domains,
Identify learning objectives of real interest to the learner
Propose learning activities to the learner
Lead the learner around the learning environment

Skills and knowledge level detection: competency level, confidence level (evidence-based confidence). For instance using dashboard to get an idea of learning progress,… and what type of skills are affected.

Teachers need to feel that it does not affect their workload, they become more open to these new options.


Question from my end on making learning visible: do you have examples of feedback from the learner that make the actual learning visible.  You mention on how learners learn, but it seems you are more viewing it from a teacher viewpoint, awareness in the learner.
Answer: analytics are coming from a variety of sources and at Austin, Texas, we also work with Codex and MI-dash see the learner progress over time, SCRL which uses self-evaluation, learning initiative design… 

No comments:

Post a Comment