computational trust
TRANSCRIPT
-
7/26/2019 Computational Trust
1/52
Joana Urbano A ril 2013 1
Computational Trust
Joana Urbano
-
7/26/2019 Computational Trust
2/52
Joana Urbano A ril 2013 2
Trust in ANTE
-
7/26/2019 Computational Trust
3/52
Joana Urbano A ril 2013 3
Why Trust?
Social and economic relationships are uncertainand bring
vulnerabilityto the interacting partners.
Trust allows to reduce the complexity of relationships.
Following the sociological view of trust, trust allows social life
(Luhmann, 1979).
-
7/26/2019 Computational Trust
4/52
Joana Urbano A ril 2013 4
Why Computational Trust?
Enabler technology for the automatization & virtualizationof
business:
E-procurement (B2B).
E-commerce & auctions (B2C, C2C, C2B).
Management of distributed resources (grid, cloud).
Desired feature of social networks.
It has received attention from national and EU funding: eRep, eTrustCom, COST Action IC0801 (WG5), LIACC FCT project,
-
7/26/2019 Computational Trust
5/52
Joana Urbano A ril 2013 5
What is Trust?
Still no definition accepted by all.
Probably the most consensual definition:
Trust is both apositive expectationthat the trustee will act as expected,
and a willingness or intention to accept vulnerability(Mayer:1995).
Some well know properties: It depends on the evaluator (I trust X as a vendor but Y does not).
It depends on the role(I trust X as a vendor but not as a buyer).
It depends on the context(I trust X as an exporter of cotton to Europe but not as anexporter of chiffon to the USA).
It depends on time(I used to trust X, but i do not trust it anymore).
It does not depend necessarily on reputation(I trust X despite its bad reputation).
-
7/26/2019 Computational Trust
6/52
Joana Urbano A ril 2013 6
Different Approaches to Trust
Psychology:trust is considered at the individual level, as a personality
trait (questionnaire surveys, other psychometric scaling techniques, trust games).
Social psychology and Sociology:trust is understood as a property of
ongoing dyads, groups, and collectivities.
Neuroscience:activation of the amygdala in kinship relationships, brain
damage and trust, etc.
Economy:how trust is affected by incentives, norms, institutions, and
other governance mechanisms.
Management and Organizations:the relation between trust, power,
opportunism and betrayal, amongst others.
-
7/26/2019 Computational Trust
7/52
Joana Urbano A ril 2013 7
The Computational Trust Approach
Question:how computer scientists deal with all this
diversity of trust concepts and approaches?
Answer:very badly!
Real implementations are too simplistic (Amazon.com,
eBay.com, epinions.com, etc.).
Research on computational trust tend to forget the social
nature of trust.
-
7/26/2019 Computational Trust
8/52
Joana Urbano A ril 2013 8
Agenda
In this talk:
1. Simple computational trust algorithms in e-commerce.
2. Overview of research on computational trust.
3. New trends on computational trust.
4. LIACCs current approach.
-
7/26/2019 Computational Trust
9/52
Joana Urbano A ril 2013 9
Simple Trust Systems in the Internet
Amazon.com (B2C), eBay.com (C2C)
-
7/26/2019 Computational Trust
10/52
Joana Urbano A ril 2013 10
Amazon.comTrust in Products
five star classification
-
7/26/2019 Computational Trust
11/52Joana Urbano A ril 2013 11
Amazon.comCredibility of Reviewers
-
7/26/2019 Computational Trust
12/52Joana Urbano A ril 2013 12
Amazon.comReviewers Rank
A reviewer's rankis determined by the overall helpfulness of all their reviews,
factoring in the number of reviews they have written. More weight is given to
recent reviews.
-
7/26/2019 Computational Trust
13/52Joana Urbano A ril 2013 13
Amazon.comTrust in Sellers
-
7/26/2019 Computational Trust
14/52Joana Urbano A ril 2013 14
Amazon.comTrust in Sellers
Amazon TipsLook for clues within a
sellers feedback that
indicate the following:
the ability to fulfil
and ship orders in a
timely fashion. a sellers willingness
to resolve
transaction disputes.
an indication that
the quality of the
products shipped
matches the
descriptionsupplied
by the seller.
-
7/26/2019 Computational Trust
15/52Joana Urbano A ril 2013 15
eBay.comTrust in Sellers
Trust on Sellersadd detailed seller rankings, contact of seller, web of favorite
sellers.
-
7/26/2019 Computational Trust
16/52Joana Urbano A ril 2013 16
eBay.comTop-Rated Sellers
-
7/26/2019 Computational Trust
17/52Joana Urbano A ril 2013 17
eBay.comTrust in the System
-
7/26/2019 Computational Trust
18/52Joana Urbano A ril 2013 18
Considerations About Trust in Amazon & eBay
Trust measures are simple or weighed means.
Much is left to comments written in natural language (e.g.,
delivery, quality, resolution of disputes)context is not
addressed in an automatic way.
Long-term relationships are not the pattern.
eBay foruns reportretalliation cases.
A few negative classifications removea seller from business.
-
7/26/2019 Computational Trust
19/52Joana Urbano A ril 2013 19
Research on Computational Trust
Traditional Approaches
-
7/26/2019 Computational Trust
20/52Joana Urbano A ril 2013 20
Traditional Approaches to Computational Trust
Trust is a Cognitive Construct: it builds on existing
past evidence about the trustees behavior
1. Alice reported that Bob was a very good tour guide when
they visited Trondheim at night
2. Jim reported that Bob was a lousy tour guide when they
visited Bergen under stormy weather
-
7/26/2019 Computational Trust
21/52Joana Urbano A ril 2013 21
Traditional Approaches to Computational Trust
Trust is about Trustworthiness
I trust Bob as a tour guide under normal conditions
because a lot of people (and probably myself) reported
that he is trustworthyunder these conditions
Trust is Contextual
I trust Bob as a tour guide under normal conditions but I
do not trust him under stormy weather
-
7/26/2019 Computational Trust
22/52Joana Urbano A ril 2013 22
Trustworthiness Computation
How to compute the trustworthiness of this agent?
fabric quantity del. time outcome
cotton medium big fulfill
cotton high medium fulfill
cotton medium big fulfill
voile medium medium fulfill
voile high low violate
cotton high medium fulfillvoile medium big fulfill
voile high big fulfill
cotton low big fulfill
voile medium medium fulfill
voile medium big fulfill
voile medium medium fulfillcotton low big fulfill
cotton high medium violate
voile medium low fulfill
voile medium big fulfill
voile medium low fulfill
-
7/26/2019 Computational Trust
23/52Joana Urbano A ril 2013 23
Trustworthiness Computation
Weighted Means by Recency
Trustworthiness of trustee is given by aggregating evidence using
weighted means by recency:
Model: Fire (Huynh, Jennings, Shadbolt, 2006)
)5.0ln(,,
.
)(1
0
1
0 dew
w
vw
NTit
iN
i
i
i
N
i
i
-
7/26/2019 Computational Trust
24/52Joana Urbano A ril 2013 24
Weighted Means By Recency
d = 5, t = 16
lambda = -5/log(0.5)
wi = exp (-delta/lambda)
fabric quantity del. time outcome v t delta w w*v
cotton medium big fulfill 1 0 16 0,38 0,38
cotton high medium fulfill 1 1 15 0,41 0,41
cotton medium big fulfill 1 2 14 0,43 0,43
voile medium medium fulfill 1 3 13 0,46 0,46
voile high low violate -1 4 12 0,49 -0,49cotton high medium fulfill 1 5 11 0,52 0,52
voile medium big fulfill 1 6 10 0,55 0,55
voile high big fulfill 1 7 9 0,58 0,58
cotton low big fulfill 1 8 8 0,62 0,62
voile medium medium fulfill 1 9 7 0,66 0,66
voile medium big fulfill 1 10 6 0,70 0,70
voile medium medium fulfill 1 11 5 0,74 0,74
cotton low big fulfill 1 12 4 0,79 0,79
cotton high medium violate -1 13 3 0,83 -0,83
voile medium low fulfill 1 14 2 0,89 0,89
voile medium big fulfill 1 15 1 0,94 0,94
voile medium low fulfill 1 16 0 1,00 1,00
10,96 8,32
tw = 0,76
)5.0ln(,,
.
)(1
0
1
0 dew
w
vw
NTit
iN
ii
i
N
i
i
h
-
7/26/2019 Computational Trust
25/52Joana Urbano A ril 2013 25
Trustworthiness Computation
Beta Reputation System Trustworthiness and
reputation are modeled as
the expectation of a beta
PDF (Josang et al., 2002)
http://en.wikipedia.org/wiki/File:Beta_distribution_pdf.png
i
-
7/26/2019 Computational Trust
26/52Joana Urbano A ril 2013 26
Beta Reputation System
lambda = 0,9; delta = 16
r_total = r_total + r_opinion *pow(lambda, delta-i)
fabric quantity del. time outcome r_opinion s_opinion t r_total s_total
cotton medium big fulfill 1,0 0,0 0 0,185 0,000
cotton high medium fulfill 1,0 0,0 1 0,391 0,000
cotton medium big fulfill 1,0 0,0 2 0,620 0,000
voile medium medium fulfill 1,0 0,0 3 0,874 0,000
voile high low violate 0,0 1,0 4 0,874 0,282
cotton high medium fulfill 1,0 0,0 5 1,188 0,282voile medium big fulfill 1,0 0,0 6 1,537 0,282
voile high big fulfill 1,0 0,0 7 1,924 0,282
cotton low big fulfill 1,0 0,0 8 2,355 0,282
voile medium medium fulfill 1,0 0,0 9 2,833 0,282
voile medium big fulfill 1,0 0,0 10 3,364 0,282
voile medium medium fulfill 1,0 0,0 11 3,955 0,282
cotton low big fulfill 1,0 0,0 12 4,611 0,282cotton high medium violate 0,0 1,0 13 4,611 1,011
voile medium low fulfill 1,0 0,0 14 5,421 1,011
voile medium big fulfill 1,0 0,0 15 6,321 1,011
voile medium low fulfill 1,0 0,0 16 7,321 1,011
48,38 6,59
tw = 0,87
T hi C i
-
7/26/2019 Computational Trust
27/52Joana Urbano A ril 2013 27
Trustworthiness Computation
Weighted Means by Similarity of Situation
Situation is represented as point in a multi-dimensional context space
Model: CSRC (Rehak et al. 2008)
)2,1(
1
0
1
0 ,
.
)( ccdiN
i
i
i
N
i
i
ew
w
vw
NT
W i h d M B Si il i f Si i
-
7/26/2019 Computational Trust
28/52Joana Urbano A ril 2013 28
Weighted Means By Similarity of Situation
fabric quantity del. time outcome
cotton medium big fulfill
cotton high medium fulfill
cotton medium big fulfill
voile medium medium fulfill
voile high low violate
cotton high medium fulfill
voile medium big fulfill
voile high big fulfillcotton low big fulfill
voile medium medium fulfill
voile medium big fulfill
voile medium medium fulfill
cotton low big fulfill
cotton high medium violate
voile medium low fulfillvoile medium big fulfill
voile medium low fulfill
)2,1(
1
0
1
0 ,
.
)( ccdiN
i
i
i
N
i
i
ew
w
vw
NT
W i h d M B Si il i f Si i
-
7/26/2019 Computational Trust
29/52Joana Urbano A ril 2013 29
Weighted Means By Similarity of Situation
It may be hard to tune the distance functions!
T t thi C t ti
-
7/26/2019 Computational Trust
30/52
Joana Urbano A ril 2013 30
)(
||
||)(),(
)(
AValuesv
vv SEntropy
S
SSEntropyASGain
Trustworthiness Computation
low
fulfill
quantity
medium high
dtime
low medium big
violate fulfill fulfill
fulfill
low
fulfill
quantity
medium high
dtime
low medium big
violate fulfill fulfill
fulfill dtime
high
low
Failure tendency: (*, high, low)
Contextual FitnessLIACCs approach to situation-aware trust.
Model: Contextual Fitness (Urbano, Rocha, Oliveira, 2010)
fabric quantity del. time outcome
cotton medium big fulfill
cotton high medium fulfill
cotton medium big fulfill
voile medium medium fulfill
voile high low violate
cotton high medium fulfill
voile medium big fulfill
voile high big fulfillcotton low big fulfill
voile medium medium fulfill
voile medium big fulfill
voile medium medium fulfill
cotton low big fulfill
cotton high medium violate
voile medium low fulfillvoile medium big fulfill
voile medium low fulfill
T t thi C t ti
-
7/26/2019 Computational Trust
31/52
Joana Urbano A ril 2013 31
Trustworthiness Computation
Other approaches:
Dirichlet distributions: beta distributions for eachdimension of evidence and correlation of dimensions.
Nave Bayesian networks: P(Y|X) = (P(X|Y).P(Y))/P(X).
LIACCs approach: Sinalpha:
Model: Sinalpha (Urbano, Rocha, Oliveira, 2009)
U i R t ti l I f ti
-
7/26/2019 Computational Trust
32/52
Joana Urbano A ril 2013 32
Bob has a good reputation as a tour guide
Using Reputational Information
Is he?
Who said that?Are these people credible?
Do they believe it?
Is Bob a good tour guide or a good guy?
etc ..
C t ti l R t ti l
-
7/26/2019 Computational Trust
33/52
Joana Urbano A ril 2013 33
Computational Reputational
Issues on computational reputation:
Credibility of sources.
Credibility of transmitted information.
Relevance of transmitted information.
Heterogeneity (both syntactic and semantic) of thereputation score being transmitted.
Incentives to referrals.
Bad mouthing.
-
7/26/2019 Computational Trust
34/52
Joana Urbano A ril 2013 34
Research on Computational Trust
New Trends
N A h A N d d!
-
7/26/2019 Computational Trust
35/52
Joana Urbano A ril 2013 35
New Approaches Are Needed!
Traditional approaches fail to correctly model trust:
1. Trust is a socio-cognitive concept.
2. Trustworthiness is not single dimensional.
3. Trust is more than trustworthiness.
4. Approaches are evaluated using extremely simple models
of agents behavior.
Conceptual Model of Mayer et al (1995)
-
7/26/2019 Computational Trust
36/52
Joana Urbano A ril 2013 36
Conceptual Model of Mayer et al. (1995)
Trust is more than trustworthiness, and trustworthiness is
multi-dimensional.
Computational Model of Kelton et al (2008)
-
7/26/2019 Computational Trust
37/52
Joana Urbano A ril 2013 37
Computational Model of Kelton et al. (2008)
Identification, context and reputation are added.
Socio Cognitive Model of Trust
-
7/26/2019 Computational Trust
38/52
Joana Urbano A ril 2013 38
Socio-Cognitive Model of Trust
A true computational trust model using Fuzzy Cognitive Maps,
but not entirely implemented (Castelfranchi & Falcone, 2010)
-
7/26/2019 Computational Trust
39/52
Joana Urbano A ril 2013 39
Research on Computational Trust
Our Approach
The SOLUM Framework
-
7/26/2019 Computational Trust
40/52
Joana Urbano A ril 2013 40
The SOLUM Framework
The Ability Evaluation Function
-
7/26/2019 Computational Trust
41/52
Joana Urbano A ril 2013 41
The Ability Evaluation Function
is obtained with Contextual Fitness
Results:
The Benevolence Evaluation Function
-
7/26/2019 Computational Trust
42/52
Joana Urbano A ril 2013 42
The Benevolence Evaluation Function
We claim that our model is truly social!
We analyze the past evidence in light of the relationshipsexisting between the trustees and the trusters
Different benevolence stages
or benevolence relationshipsmay explain the different
outcomes.
fabric quantity del. time outcome
cotton medium big fulfill
cotton high medium fulfill
cotton medium big fulfill
voile medium medium fulfill
voile high low violate
cotton high medium fulfill
voile medium big fulfill
voile high big fulfill
cotton low big fulfill
voile medium medium fulfillvoile medium big fulfill
voile medium medium fulfill
cotton low big fulfill
cotton high medium violate
voile medium low fulfill
voile medium big fulfill
voile medium low fulfill
The Social Tuner Component
-
7/26/2019 Computational Trust
43/52
Joana Urbano A ril 2013 43
The Social Tuner Component
1. Get Coefficient of Benevolent Actions
2. Calculate the Trustees Benevolence
We model benevolence as the trustees attachment toward the truster
The Social Tuner Component
-
7/26/2019 Computational Trust
44/52
Joana Urbano A ril 2013 44
The Social Tuner Component
Trustworthiness evaluation function that understands the
stage of the relationship:
benevolence
ability
The Social Tuner Component
-
7/26/2019 Computational Trust
45/52
Joana Urbano A ril 2013 45
The Social Tuner Component
So far, the results are very promising!
The Integrity Tuner
-
7/26/2019 Computational Trust
46/52
Joana Urbano A ril 2013 46
The Integrity Tuner
Get Coefficient of Integrity
Calculate the trustees Integrity
We model integrity as the consistency of past actions and the accordance to what
was established
Evaluation of Our Approach
-
7/26/2019 Computational Trust
47/52
Joana Urbano A ril 2013 47
Evaluation of Our Approach
We proposed and formalizeddifferent models of partners behavior based
on empirical workin different areas of research.
These models considered different characteristics of trusters and trustees:
Dispositional and mutualistic benevolence, ability, integrity, value of
agreement, satisfaction with partner, value of betrayal,
Publications
-
7/26/2019 Computational Trust
48/52
Joana Urbano A ril 2013 48
Publications
A. International Journals:
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira, An Approach to Computational Social Trust,
AICommunications, IOS Press, 2013, accepted February 2013.
Joana Urbano, Henrique Lopes Cardoso, Ana Paula Rocha, Eugnio Oliveira (2012), Trust and Normative
Control in Multi-Agent Systems,Advances in Distributed Computing and Artificial Intelligence Journal, Vol. I,
No. 1, July 2012, pp. 43-52.
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira, A Situation-Aware Computational Trust Model for
Selecting Partners, in N.T. Nguyen (Ed.), Transactions on Computational Collective Intelligence V, LNCS 6910,
pp. 84-105, 2011, Springer-Verlag Berlin Heidelberg 2011.
B. Book chapters:
Henrique Lopes Cardoso, Joana Urbano, Ana Paula Rocha, Antnio J. M. Castro, Eugnio Oliveira (2013).
Chap. 32: ANTE: Agreement Negotiation in Normative and Trust-enabled Environments, in S. Ossowski
(ed.),Agreement Technologies, Law, Governance and Technology Series, Volume 8, Springer, ISBN 978-94-
007-5582-6, pp. 549-564.
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira (2013). Chap. 23: A Socio-Cognitive Perspective of Trust,
in S. Ossowski (ed.),Agreement Technologies, Law, Governance and Technology Series, Volume 8, Springer,ISBN 978-94-007-5582-6, pp. 419-429.
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira, Extracting Trustworthiness Tendencies Using the
Frequency Increase Metric, Enterprise Information Systems, LNBIP, 2011, Volume 73, Part 3, 208-221
Joana Urbano, Ana Paula Rocha and Eugnio Oliveira. "Trust Evaluation for Reliable Electronic Transactions
between Business Partners", in Lecture Notes in Business Information Processing, 2012, Volume 98, Part 3,
219-237, DOI: 10.1007/978-3-642-28563-9_12
Publications
http://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/ADCAIJ12.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/ADCAIJ12.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/tcci-69100084.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/tcci-69100084.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/ICEIS-LNBIP-730208.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/ICEIS-LNBIP-730208.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/ICEIS-LNBIP-730208.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/ICEIS-LNBIP-730208.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/tcci-69100084.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/tcci-69100084.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/tcci-69100084.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/tcci-69100084.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/ADCAIJ12.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/ADCAIJ12.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/ADCAIJ12.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/ADCAIJ12.pdf -
7/26/2019 Computational Trust
49/52
Joana Urbano A ril 2013 49
Publications
C. International Conferences
Joana Urbano, Henrique Lopes Cardoso, Ana Paula Rocha and Eugnio Oliveira. "Trust and Normative
Control in Multi-agent Systems: An Empirical Study", in Highlights on Practical Applications of Agents and
Multi-Agent Systems, Series: Advances in Intelligent and Soft Computing, 2012, Volume 156/2012, 207-
214, DOI: 10.1007/978-3-642-28762-6_25
Henrique Lopes Cardoso, Joana Urbano, Pedro Brando, Ana Paula Rocha and Eugnio Oliveira. "ANTE:
Agreement Negotiation in Normative and Trust-Enabled Environments", in Advances on Practical
Applications of Agents and Multi-Agent Systems, Series: Advances in Intelligent and Soft Computing, 2012,
Volume 155/2012, 261-264, DOI: 10.1007/978-3-642-28786-2_33
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira, Dynamic Agents Behavior Model for Computational
Trust. L. Antunes and H.S. Pinto (Eds.): EPIA 2011, LNAI 7026, Springer-Verlag Berlin Heidelberg, pp. 536
550, 2011
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira, Trust-Based Selection of Partners, C. Huemer and T.
Setzer (Eds.): EC-Web 2011, LNBIP 85, pp. 221232, 2011. Springer-Verlag Berlin Heidelberg 2011
Joana Urbano, Henrique Lopes Cardoso, Eugnio Oliveira (2010). Making Electronic Contracting
Operational and Trustworthy, inAdvances in Artificial Intelligence Proceedings of the 12th Ibero-
American Conference on Artificial Intelligence (IBERAMIA 2010), Lecture Notes on Artificial Intelligence V.6433, Springer, ISBN 978-3-642-16951-9, pp. 264-273, Baha Blanca, Argentina, November 1-5, 2010
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira (2010). "Trustworthiness Tendency Incremental
Extraction Using Information Gain, Proceedings of the 2010 IEEE/WIC/ACM International Conference on
Web Intelligence and Intelligent Agent Technology, Toronto, Canada, 1-3 September, 2010, pp. 411-414.
Publications
http://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/EPIA-70260536.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/EPIA-70260536.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/ecweb11.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/iberamia2010-159-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/iberamia2010-159-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/IAT10.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/IAT10.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/IAT10.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/IAT10.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/iberamia2010-159-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/iberamia2010-159-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/ecweb11.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/ecweb11.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/ecweb11.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/EPIA-70260536.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2011/EPIA-70260536.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected]://paginas.fe.up.pt/~niadr/PUBLICATIONS/2012/[email protected] -
7/26/2019 Computational Trust
50/52
Joana Urbano A ril 2013 50
Publications
Joana Urbano, Ana Paula Rocha and Eugnio Oliveira (2010), "In the Search of Better Deals Using Trust",
proceedings of ECCAI'10 Workshop IAT4EB, pp.25-30, July, Lisboa, Portugal.
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira (2010). Refining the Trustworthiness Assessment of
Suppliers through Extraction of Stereotypes", 12th International Conference on Enterprise Information
Systems (ICEIS 2010), Funchal, Portugal, 8-12 June 2010, pp. 85-92. BEST PAPER AWARD inTrack "Artificial
Intelligence and Decision Support Systems.
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira (2010). Trust Estimation Using Contextual Fitness", 4th
International KES Symposium on Agents and Multi-agent SystemsTechnologies and Applications, Gdynia,
Poland 23 - 25 June 2010, LNAI 6070, pp.42-51
Agnieszka Danek, Joana Urbano, Ana Paula Rocha, Eugnio Oliveira (2010). Engaging the Dynamics of Trustin Computational Trust and Reputation Systems", 4th International KES Symposium on Agents and Multi-
agent SystemsTechnologies and Applications, Gdynia, Poland 23 - 25 June 2010, LNAI 6070, pp. 22-31
Joana Urbano, Ana Paula Rocha, Eugnio Oliveira, "A Trust Aggregation Engine that Uses Contextual
Information", EUMAS 2010- 7th European Workshop on Multi-Agent Systems, 12pp., Ayia Napa, Cyprus,
December 2009.
Joana Urbano, Ana Paula Rocha and Eugnio Oliveira. "Computing Confidence Values: Does Trust Dynamics
Matter?", in L. Seabra Lopes, N. Lau, P. Mariano & L. M. Rocha (eds.), Progress in Artificial Intelligence
14thPortuguese Conference on Artificial Intelligence (EPIA 2009), Springer, ISBN 978-3-642-04685-8, pp. 520-531,
Aveiro, Portugal, October 12-15, 2009.
Joana Urbano, Ana Paula Rocha, and Eugenio Oliveira (2009) "Trust Evaluation for Reliable Electronic
Transactions between Business Partners" in Klaus Fischer, Jorg P. Muller, James Odell, Arne J. Berre (eds.),
Proceedings of The AAMAS09 Workshop on Agent-based Technologies and applications for enterprise
interOPerability, pp. 85-96, Budapest, Hungary, May 12, 2009.
People Working on Trust at LIACC
http://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/ECAI2010_IAT4EB_ws.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/ICEIS10_257.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/ICEIS10_257.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/ams10-103-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/KES-AMSTA10-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/KES-AMSTA10-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/EUMAS09-59-Final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/EUMAS09-59-Final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/trustMasta09.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/trustMasta09.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/ATOP09_Joana.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/ATOP09_Joana.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/ATOP09_Joana.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/ATOP09_Joana.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/trustMasta09.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/trustMasta09.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/EUMAS09-59-Final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2009/EUMAS09-59-Final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/KES-AMSTA10-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/KES-AMSTA10-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/ams10-103-final.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/ICEIS10_257.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/ICEIS10_257.pdfhttp://paginas.fe.up.pt/~niadr/PUBLICATIONS/2010/ECAI2010_IAT4EB_ws.pdf -
7/26/2019 Computational Trust
51/52
Joana Urbano A ril 2013 51
People Working on Trust at LIACC
Active:
Joana Urbano (PhD student)
Ana Paula Rocha and Eugnio Oliveira (Coordinators)
Henrique Lopes Cardoso (trust & norms)
Past collaboration:
Agnieszka Danek (3 month grant)
Filipe Silva (MIEIC dissertation)
Srgio Moura (MIEIC dissertation)
-
7/26/2019 Computational Trust
52/52
Obrigada!