1. RECOGNITION: Relevance and
RECOGNITION: Relevance and
Cognition for Self‐Awareness in
a Content‐Centric Internet
Stuart M. Allen, Franco Bagnoli, Gualtiero Colombo,
g
Marco Conti, Jon Crowcroft, Chris Jones, Pietro Liò,
Refik Molva, Melek Onen, Andrea Passarella,
Ioannis Stavrakakis, Roger M. Whitaker, Eiko Yoneki
k k h k k k
RECOGNITION overview
1
December 2010
2. Motivation: Technological Trends
Motivation: Technological Trends
• Participatory generation of content
p yg
– Prosumers, diversity, expanding edges
– Long tail, swamping, scale!
• Content in the environment
– Linkage of the physical and virtual worlds
– Embedding content and knowledge
• Acquiring knowledge through social
q g g g
mechanisms
– Blogging, social networking,
recommendation, RSS feeds…
• How content reaches users will
continue to change…
ti t h
RECOGNITION overview
2
December 2010
3. Self‐awareness to support
technological trends
• Our Intention: Paradigm to support
ICT functions
ICT f ti
– Enabling content centricity
• Better fitting of users to content and vice
Better fitting of users to content and vice
versa
– Synchronize content with human activity
and needs
• Place, time, situation, relevance, context,
social search
social search
– Autonomic management
• Of content, its acquisition and resource
utilization
l
RECOGNITION overview
3
December 2010
4. Human Awareness Behaviours
Human Awareness Behaviours
• A
Approach: Capture & exploit key
h C & l i k
behaviours of the most intelligent
living species
living species
– Human capability is phenomenal in
navigating complex & diverse stimuli
navigating complex & diverse stimuli
– Filter & suppress information in “noisy”
situations with ambient stimuli
– Extract knowledge in presence of
uncertainty
–EExercise rapid value judgment for
i id l j d tf
prioritisation
– Engage a social context and multi‐scale
Engage a social context and multi scale
learning RECOGNITION overview
4
December 2010
5. Human Awareness Behaviours
Human Awareness Behaviours
Cognitive psychological basis
For awareness and understanding
Defining key principles for exploitation by
technology components
technology components
Embedding these principles for
self‐awareness in autonomic content
acquisition in pervasive environments
Potential change in behaviour due to
self–awareness in ICT
RECOGNITION overview
5
December 2010
6. Overview of Structure
Overview of Structure
UNIFI LEAD
CU LEAD
CNR LEAD
NKUA LEAD UCAM LEAD
RECOGNITION overview
6
December 2010
7. Providing Autonomic Content
Management
• Th
Through Recognition “Nodes”, content becomes as self‐
hR iti “N d ” t tb lf
aware as devices
• Allow individuals to gain content that they didn’t know
g y
they wanted…
• Geo‐Informatics: space, place, time…
– C t t l
Content placement & retrieval based on situation and location
t& t i lb d it ti d l ti
• Storage and forwarding decisions based on relevance from:
– Social context
Social context
– Location & environment
• Trust & security management
– Uncertainty & belief
RECOGNITION overview
7
December 2010
8. Interdisciplinary Dimensions
Interdisciplinary Dimensions
– Complex systems
– Artificial intelligence
– Geo‐informatics
– Cognitive psychology
Cognitive psychology
– Information retrieval
– Communication systems
– Security, trust
RECOGNITION overview
8
December 2010
9. Key Questions…
Key Questions
• Psychology
– What key concepts should be develop/include?
y p p/
– Can these be used in different parts of the project?
• Scenarios
– What contemporary areas of “social computing” are
key to prioritise?
key to prioritise?
– What would have the biggest impact?
– Are there demo’s that could be developed?
• Other questions…….
RECOGNITION overview
9
December 2010
10. Proposal: Psychology areas
Proposal: “Psychology” areas
• Recognition, Probabilistic Mental Models,
b bl l d l
Heuristics
–HHuman characteristics for agents
h t i ti f t
– Decision making under bounded rationality
• Social Learning
Social Learning
– Observing, retaining, learning, replicating (mimicking)
• Spatial Cognition
Spatial Cognition
– Space, place, context
• Belief Desire and Intention models
Belief, Desire and Intention models
– Pulling from different areas of psychology but not fully
grounded
RECOGNITION overview
10
December 2010
11. 1 ‐ Relevance Theory
1 Relevance Theory
• Sperber and Wilson
p
– Non‐coding model of communication
– Inferential model taking into account
context via “utterances”
– provide "cognitive effects" worthy of the
processing effort required to find the
processing effort required to find the
meaning
• The speaker purposefully gives a clue to the
hearer
• The hearer infers the intention from the clue
and the context‐mediated information. The
hearer must interpret the clue, taking into
account the context, and surmise what the
speaker intended to communicate.
RECOGNITION overview
11
December 2010
12. 2‐ Judgment & Decision Making
2 Judgment & Decision Making
• Work of Daniel Goldstein et al
– Heuristics that make us smart…
• “Take the best” heuristic
• Recognition heuristic
– Bounded rationality
Bounded rationality
• Limited direct knowledge/partial info
• Fast inference has to be made
Fast inference has to be made…
RECOGNITION overview
12
December 2010
13. 2‐ Judgment & Decision Making
2 Judgment & Decision Making
• Take the best heuristic
Take the best heuristic
– judgment based on multiple criteria
• the criteria are tried one at a time
the criteria are tried one at a time
according to their “cue validity”
• high cue validity for a given feature
g y g
means that the feature or attribute is
more diagnostic of the class membership
than a feature with low cue validity
than a feature with low cue validity
– a decision is made based on the first
discriminating criterion
discriminating criterion
• the heuristic did well at making accurate
inferences in real world environments
RECOGNITION overview
13
December 2010
14. 2‐ Judgment & Decision Making
2 Judgment & Decision Making
• Recognition heuristic
Recognition heuristic
– If one of two objects is recognized and
the other is not, then infer that the
the other is not then infer that the
recognized object has the higher value
with respect to the criterion.
p
– Sensitive to the criterion
• Methodology for “cue validity”
Methodology for cue validity
– Less‐is‐more effect
• Limited information does not impede
Limited information does not impede
performance (to the contrary!)
RECOGNITION overview
14
December 2010
15. 3‐ Spatial Cognition
3 Spatial Cognition
• Human understanding and meaning for
Human understanding and meaning for
ill‐defined but commonly used spatial
terms
• South east…
• South Wales
• Central london
• Use of these in geo‐spatial content
g p
so that it can become self‐aware
RECOGNITION overview
15
December 2010
16. Key Questions…
Key Questions
• Psychology
– What key concepts should be develop/include?
y p p/
– Can these be used in different parts of the project?
• Scenarios
– What contemporary areas of “social computing” are
key to prioritise?
key to prioritise?
– What would have the biggest impact?
– Are there demo’s that could be developed?
• Other questions…….
RECOGNITION overview
16
December 2010
17. Candidate Scenarios
Candidate Scenarios
• Information Retrieval & content provision
– Human awareness when using search engine
interfaces – e.g., automatic cue detection & HCI
• Self‐aware Multimedia and “Active” Data
– MP3, other types of content
– Self‐aware meta‐data for spatial problems
Self‐aware meta‐data for spatial problems
• Social Computing
– Crowd sourcing, recommendation, filtering, micro‐
d i d i fil i i
blogging, tagging
RECOGNITION overview
17
December 2010
18. RECOGNITION: Relevance and
RECOGNITION: Relevance and
Cognition for Self‐Awareness in
a Content‐Centric Internet
Stuart M. Allen, Franco Bagnoli, Gualtiero Colombo,
g
Marco Conti, Jon Crowcroft, Chris Jones, Pietro Liò,
Refik Molva, Melek Onen, Andrea Passarella,
Ioannis Stavrakakis, Roger M. Whitaker, Eiko Yoneki
k k h k k k
RECOGNITION overview
18
December 2010