Transcript CoWiSe
“Computational Wisdom and Self-Computing”
research group objectives
Vagan Terziyan
Faculty of Information Technology, University of Jyvaskyla, Finland
Jyvaskyla, 6 October 2016
Smart-Resource: Agent-driven Predictive and Preventive
Maintenance
operator
field crew
expert
consumers owner
manager
administration
3
Smart-Resource: Agent-driven Predictive and Preventive
Maintenance (“personal semantic health record” for things)
Sensors and
alarm detectors
Resource
info
Operators
Experts
Software and
services
AI tools
(Knowledge
Discovery)
Maintenance
workers
Other users
4
From Smart-Resource and UBIWARE to Smart Health: Agent-driven
Predictive and Preventive Healthcare system
2
Online
Monitoring
Sensing
Testing
Diagnostics
Treatment
4
3
1
5
The “Battle” of the New Millennium
If we want to survive with the Big Data, then we
must allow it to be autonomous and self-managed
The toolset of
“collective intelligence”
must be “in the hands”
of Big Data itself as
autonomous and selfmanaged entity
Humans
Mind “clones” of humans
Machine-Learningdriven agents
The technology behind Cognitive Computing
relies on advances in the study of Collective
Intelligence, in regards to not only physical
groups of humans, but more to the conceptual
and mechanical
systems we build.
Humans
Mind “clones” of humans
Machine-Learningdriven agents
Cognitive Computing is
the simulation of human
thought processes in a
computerized model.
Cognitive computing
involves self-learning
systems that use data
mining, pattern
recognition and natural
language processing to
mimic the way the
human brain works. The
goal of cognitive
computing is to automate
decision-making and
problem-solving.
“All you need is
WISDOM” !
WISDOM
LONG list of
alternatives for the
decision and relevant
input BIG data
Decision made /
chosen alternative
“Wise” decision-making includes realizing lack of resources for the optimal decision due to big data to
be processed, finding compromise between efficiency and effectiveness of the potential decision and
smart utilization of the instrument (focusing-filtering-forgetting-contextualizing-compressingconnecting) for giving-up something yet making reasonable decision (“wise decision”).
WISDOM-I: Capability to compromise between the efficiency and the
effectiveness when addressing the Big Data challenge
Efficiency is achieved if the ratio of the
effort (resource) spent is reasonable
comparably to the utility of the result.
E.g., if a result is not timely the utility of
the resulting knowledge will drop.
Effectiveness is achieved if:
(a) not a single important data/knowledge
token is left unattended (completeness); and
(b) these tokens are processed adequately for
further consumption (expressiveness/granularity).
WISDOM-I tools:
Ermolayev V., Akerkar R., Terziyan V., Cochez M., Towards Evolving Knowledge Ecosystems for Big
Data Understanding, In: R. Akerkar (ed.), Big Data Computing, Chapman and Hall, 2014 (Ch. 1, pp. 3-55).
WISDOM-II: Capability to balance between evolving configuration and
challenges of the external environment and own (internal)
configuration and objectives
WISDOM-II tool: Self-Computing/Self-Management
Terziyan V., Challenges of the “Global Understanding Environment” based on Agent Mobility,
In: V. Sugumaran (ed.), Application of Agents and Intelligent Information Technologies, IGI, 2007,
(Ch. 7, pp.121-152).
WISDOM-III: Capability to decide “ethically” and “emotionally” correct
WISDOM-III tool: Computational Culture and Ethical Computing
Terziyan V., Kaikova O., The "Magic Square": A Roadmap towards Emotional Business
Intelligence, Journal of Decision Systems, Vol. 24, No.3, 2015, Taylor & Francis, pp. 255-272.
WISDOM-IV: Capability to utilize “human-clones” (“mind-robots” of
human decision-makers)
WISDOM-IV tool: Pi-Mind (“Patented Intelligence”)
π (PI) – “Patented Intelligence”, with the meaning of
formalizing, licensing, sharing, reuse and integration of the
personal wisdom&value-driven decision culture for the quality,
transparency and automation of the decision-driven processes
in autonomous cyber-physical systems dealing with big data.
π -mind (patented mind): a digital patented copy of a human`s decision system providing formalization of his or her wisdom, values
system and a decision scheme used for a specific task. π-mind helps keeping and sharing an explicit ontological model of a human`s
wisdom and value system for its further use by users of the ecosystem. π-mind characterizes deliberate and formalized rules used by a
person for wise decision making in situations defined by a state of the environment to achieve specific goals. Usually in decision support
systems it`s a knowledge base which stores expert knowledge in some domain but we propose a more subjective entity: wisdom&value
base which stores specific opinions of a concrete person about the importance of various things and phenomena which he/she uses during
decision making - so called π-mind. Such space is a complex structure - non-linear and multidimensional.
Terziyan V., Golovianko M., Shevchenko O., Semantic Portal as a Tool for Structural Reform of the
Ukrainian Educational System, In: Information Technology for Development, Vol. 21, No. 3, 2015,
Taylor & Francis, pp. 381-402. See also: http://www.mit.jyu.fi/ai/Quality-3_en.pptx
WISDOM-V: Learning Wisdom
WISDOM-V tool: “Agile” Deep Learning and Wisdom Discovery
Traditional Machine Learning Process (“Learning Intelligence”)
Traditional Machine Learning Process (“Learning Intelligence”)
“Agile” Machine Learning Process (“Learning Wisdom”)
I
II
III
N
I
II
III
N
Good Engineers ...
… converting problems into products
Good Engineers vs. Managers
… “talking” about products
to make value out of it
Good Engineers vs. VIP
Engineers
… inventing new problems
VIP Engineers vs. WISE
Engineers
… inventing new and SMART problems
… designing SMART problem solvers!
Train your skills with us
Our other research
ambitions:
FROM ("SCIENTIFIC COMPUTING" OR "COMPUTATIONAL
SCIENCE") TO ("SELF-COMPUTING") ;
FROM ("ARTIFICIAL INTELLIGENCE") TO ("ARTIFICIAL WISDOM") ;
FROM ("COMPUTATIONAL INTELLIGENCE") TO ("COMPUTATIONAL
WISDOM") ;
FROM ("COGNITIVE SCIENCE") TO ("SELF-COGNITION") ;
FROM ("DATA MINING AND KNOWLEDGE DISCOVERY") TO ("SELFMINING AND WISDOM DISCOVERY");
FROM ("KNOWLEDGE MANAGEMENT") TO ("WISDOM MANAGEMENT");
FROM ("INTERNET-OF-THINGS") TO ("INTERNET-OF-WISE-THINGS");
FROM ("SMART PRODUCTS") TO ("WISE PRODUCTS");
FROM ("SMART ARCHITECTURE") TO ("WISE ARCHITECTURE");
FROM ("CYBER-SECURITY") TO ("WISE SECURITY" OR "SELFPROTECTION");
FROM ("WEB-OF-INTELLIGENCE") TO ("WEB-OF-WISDOM");
FROM ("BIG DATA") TO ("WISE DATA").
Short Summary
We consider “Computational Wisdom and SelfComputing“ to be a new paradigm of "wise" behavior of
artificial smart systems against Big Data challenge;
Artificial Intelligence and Cognitive Computing are
currently based on world cognition instruments like
machine learning, data mining and knowledge discovery;
while “Artificial Wisdom” (Self-Computing) assumes
also [deep, emotional, ethical, pragmatic, intuitive and
creative] machine-self-learning, self-mining and selfdiscovery (i.e., “self-cognition”) .
CoWiSe: “Computational Wisdom and SelfComputing” Research Group
contact: [email protected]
http://www.mit.jyu.fi/ai/vagan/index.html