Modern learning models require linking experiences in training environments with experiences in the real-world. However, data about real-world experiences is notoriously hard to collect. Social spaces bring new opportunities to tackle this challenge, supplying digital traces where people talk about their real-world experiences. These traces can become valuable resource, especially in ill-defined domains that embed multiple interpretations. The paper presents a unique approach to aggregate content from social spaces into a semantic-enriched data browser to facilitate informal learning in ill-defined domains. This work pioneers a new way to exploit digital traces about real-world experiences as authentic examples in informal learning contexts. An exploratory study is used to determine both strengths and areas needing attention. The results suggest that semantics can be successfully used in social spaces for informal learning – especially when combined with carefully designed nudges.
3. Exploitation of digital traces
as a source for informal learning
Exploration
Environment
digital
traces
digimind.com
Taming Semantic Web Technologies
•Retrieve • Semantic Data browsers
•Aggregate • Semantic Nudges
•Organise
4. Semantic Data Browser
Focus Concept
Facts
Eye Contact is Body Language
Social
Content
5. Processing Pipeline:
Semantic Data Browser
Digital Traces
Collection
Semantic
Augmentation &
Browsing &
Query Interaction
Bespoke
Ontologies Ontology
& Linked Underpinning
Data Cloud
6. Processing Pipeline:
DTs collection
•Availability of Social Web APIs
•Noise filtration mechanisms*
•Role of tutors/trainers in setting gold standard**
Digital Traces
Collection
Semantic
Augmentation &
Browsing &
Query Interaction
Bespoke
Ontologies Ontology
& Linked Underpinning
Data Cloud
* Ammari, A., Lau, L. Dimitrova, V. Deriving Group Profiles from social media, LAK 2012
** Redecker, C. et al. Learning 2.0- the impact of social media on learning in Europe, Policy Brief, European
Commission, JRC, 2010
8. Ontology Underpinning
Stage 1: Activity Modelling on Interpersonal Communication
Analysis
Activity Use Case
Theory Activity other relevant
Social Web
on a Use Model ontologies
Case
WN-
Body Affect
Stage 2: Activity Modelling Enrichment using Semantics Language
Social Multi-layered
Web Activity Modelling Ontology
(AMOn) for
Interpersonal Communications
Logical Encoding
Thakker, D. A., Dimitrova, V., Lau, L., Denaux, R., Karanasios, S., & Yang-Turner, F. (2011,). A Priori Ontology Modularisation in Ill-defined Domains.
In I-Semantics (7th International Conference on Semantic Systems). Graz, Austria
http://imash.leeds.ac.uk/ontology/amon/
9. Semantic Augmentation
Service
Purpose :
Generic service designed to link
content with the concepts from the
ontological knowledge bases in order to Handshake BL
fully benefit from the reasoning capabilities
of semantic technologies.
language
Body
Components: Handshake
• Information Extraction: Finding Simulators
mentions of entities in text reserved for when you wish to show you are in charge.
handshake is almost always best. An authority handshake should be
• Semantic Linking: between entity
mentions and ontologies, linked data
• Semantic Repository: forward chaining
repository for semantic expansion
• Ontologies: AMOn & External Information
ontologies Semantic Linking
Extraction
Implementation:
• RESTful interface for easy Semantic Ontology
integration Repository AMOn
10. Semantic Query Service
Purpose :
Generic service for querying and browsing
using semantically augmented content. In I-
CAW, it allows searching of socially and
locally authored data for real-world activities
from the domain of interest
Components: Related Content
• Concept Filtering: Identify matching Matching Content
concepts and relevant information
Term(s), Browsing
Simulators
• Content Filtering: Identify matching Concept(s) Tag Cloud
contents and relevant information
• Concept Frequency: CF/IDF analysis
• Semantic Relatedness: Content &
concept relatedness
Concept Content
Implementation: Filtering Filtering
• RESTful interface
• Contribution to the semantic browsing of
content and knowledge bases
Concept Frequency Semantic
Relatedness
11. Semantic Data Browser for Learning
• Data browsers Limitations:
• Learner is in-charge
• Cognitive onus on Learner
• Best course of action for him/her.
• Require to build on sound frameworks
• Intelligent techniques to extend I-CAW with
features that facilitate informal learning, yet
preserve the exploratory nature of social
environments
12. Nudge – Choice Architecture
•Influence options in a way that
will support choosers to act in
their own interest, preserving
freedom of choice
•This can be realized via suitable
alerts or nudges
•Nudge should alert people's
behavior in a predictable way and
at the same time it should be easy
and cheap to avoid
13. Nudges in Learning
• (Kravčík & Klamma, 2011) studied the potential impact of such
architecture on learning environments and listed
some recommendation/interpretation in the TEL context:
• For learning to be deep, it is necessary to provide various perspectives
on the same topic.
• People like stories but they may be misleading and oversimplified. Thus
it is crucial to choose suitable analogies and metaphors.
• In addition to confirming examples, also counter-examples are
important to demonstrate the validity of hypotheses and to assess risks.
Miloš Kravčík and Ralf Klamma. 2011. On psychological aspects of learning environments design. In Proceedings of the 6th European conference
on Technology enhanced learning: towards ubiquitous learning (EC-TEL'11), Carlos Delgado Kloos, Denis Gillet, Raquel M. Crespo García, Fridolin
Wild, and Martin Wolpers (Eds.). Springer-Verlag, Berlin, Heidelberg, 436-441..
14. Our contribution: Semantic
Nudges
• How some of these can be achieved using semantic web
technologies.
• We map nudges against the semantic technologies that
allows achieving them
Nudges/Choice
Architecture Impact on LE
Semantic
Nudges
15. I-CAW: introducing semantic
nudges
Digital Traces
Collection Semantic
Semantic Nudges Browsing &
Augmentation & Interaction
Activity Query
Model Ontology
Ontology
(AMOn) & Underpinning
Linked
Data Cloud
16. Semantic Nudges: Signposting
• Default Options
• usually lot of people end up with this
• The choice architecture encourages careful design of default
choices
• "default", usually lots of people will end up with it.
• In semantic Data Browsers:
• have information on a focus concept
• facts from ontological knowledge bases.
• The exact facts and amount of these facts that are available to read
while browsing affects what the learners read, the path they can
take for browsing and ultimately their awareness.
Entity Summarisation
17. Semantic Nudges: Prompts
• According to the choice architecture, the choice
architect is intended to influence the choices in
a way that will make the choosers better off, as
judged by themselves.
• This can be realized via suitable alerts
• This influence can be realised via suitable
prompts as non-invasive suggestions based on
similar and/or contradictory facts(factual
knowledge)/content.
20. Exploratory Study
Q: What are potential benefits of using semantically
augmented DTs and nudges in social spaces for informal
learning, and what are the further issues to address?
•Average age 44 years •Average age 28 years
•Experiece: 3 with 10/15 •Experiece: 3 with 1-5
Interviews, 2 with >15 Interviews, 2 with 5-10
digital
interviews interviews
traces
Interviewers Interviewee
21. Exploratory Study
What nonverbal cues can be observed in job
interview situations? (same for both groups)
What nonverbal cues show nervousness? (same
for both groups)
How would an interviewer deal with an aggressive
applicant? (for group 1) and How would an
applicant deal with an aggressive interviewer?
(for group 2)
22. Liked the authenticity of the
Liked the authenticity of the
content, content as Stimuli
content, content as Stimuli
“Examples are the beauty of system – I will learn from examples
[p10]”
“Anything that facilitates the preparation of training material and
provides real world examples to backup training is very helpful [p5]”
•Stimuli
–Further reflect on their experiences, and in some
cases help articulate what they had been doing
intuitively
–Provide their viewpoints (due to culture,
environment, tacit knowledge)
–Sense the diversity or consensus on the selected
topic
23. Content can be taken as Norm
(needs contextualising)
For example, comment
“The interviewer has his hands in front of him, which indicates that
he is concentrating and not fidgeting...”.
P5 and P10 stressed that inexperienced users may see a comment
in isolation and believe it would be valid in all situations
It was suggested that short comments could be augmented with
contextual information to assist the assessment of the credibility
of the different viewpoints
24. Semantic Nudges: Potential for
informal leaning in Semantic Browser
Signposting
•a fruitful way to provide a quick summary for
understanding a concept
•for exploration which leads to something new
to learn.
Prompts
•Task setting (pointed at aspects participants
might have not thought about )
•Complimentary knowledge (prompting
participants to “look at the task in a holistic
way”, “pointing at alternatives”, and “helping
see the big picture”)
25. Nudges need contextualising,
explicit viewpoints
•Contextualise prompts – social content
context , user interaction history, interaction
focus
•Elicit viewpoints and make it explicit
•Different prompts (e.g. complimentary)
•Different strategy for signposting
26. 1. Learning context:
– Informal Learning is important
– Social spaces and user generated content offer
new opportunities
2. Technology:
– Nudges to empower exploration
– Semantics is a promising technique for
implementing nudges (“Semantic Nudges”)
Thank You!
Dr Dhaval Thakker, Research Fellow, University of Leeds
D.Thakker@leeds.ac.uk
http://www.imreal-project.eu/
Editor's Notes
Conference 21 st century skills Tom – 21 st century Social media savvy Recent graduate So informal learning does happen for the 21 st century learners So far ad hoc bases
Just for Dhaval: If you noticed in the previous slide, what if he did not know about handshake, how does he comes to that Exploratory search seems suitable for this Or if he knew abt handshake, he can benefit from nowing what are the other body language around handshake?
Semantic data browsers[3] are the new breed of applications to come from the research efforts in the semantic community. Such browsers offer browsing of ontologies and semantically augmented data (e.g. content) by laying out browsing trajectories using relationships in the ontologies.
Semantic browsers can offer opportunities to build learning environments in which exploration of content is governed by ontologies that capture contextual aspects. Data browsers assume that the users are in charge of what they do when using the browser. This puts the cognitive onus on the user, and is particularly acute in the case of a user being a learner, i.e. not familiar with the conceptual space in the domain and may be unable to decide what is the best course of action for him/her.
Mainly used in the public administrators and policy making, but also businesses ….
And learning…in terms of suggestions…not a technical solution…
Procedure and data collection. In each session, a participant was firstly introduced to I-CAW [5 min] by following a script to perform a simple independent task. A standard script with the three tasks (Table 2) was then given to the participant which required the use of search box or signposting (All Facts, Key Facts and Overview) to find/browse relevant examples in I-CAW. When the participant finished a task, a semantic prompt was presented by the system when appropriate (e.g. task 2 included a similarity-based prompt, and task 3 included a contradiction prompt). After a participant completed all the tasks, the experimenter collected the participant’s feedback on his/her experience with I-CAW (using a semi-structured interview and a questionnaire). The materials for the study are available online. http://imash.leeds.ac.uk/imreal/icaw.html#evaluation
Two most experienced interviewers(p5 and p10) commented that some content could be mistaken as the norm .