Computing Confidence Values: Does Trust Dynamics Matter?
read more
Citations
CRM: An efficient trust and reputation model for agent computing
Partial identities as a foundation for trust and reputation
Magentix2: A privacy-enhancing Agent Platform
The Relevance of Categories for Trusting Information Sources
On the Users’ Acceptance of IoT Systems: A Theoretical Approach
References
Formalising Trust as a Computational Concept
The Beta Reputation System
An integrated trust and reputation model for open multi-agent systems
Principles of trust for MAS: cognitive anatomy, social importance, and quantification
Related Papers (5)
Frequently Asked Questions (13)
Q2. What have the authors stated for future works in "Computing confidence values: does trust dynamics matter?" ?
Namely, the authors propose as future work to identify and to categorize patterns of behaviour as new target evidences appear, through the usage of clustering techniques ; and to consider multi-attribute evaluations ( e. g. price, delivery time, and quality ).
Q3. What is the hypothesis of the paper?
Their hypothesis is that the use of an aggregation engine that encompasses the past experiences of the trustee agent and that accounts for fundamental dynamics of trust could allow for a better estimation of the trustee trustworthiness than probabilistic and statistical approaches that exist in the literature.
Q4. How many suppliers and buyers were instantiated in each experiment?
In every experiment, the authors instantiated 16 suppliers and 8 buyers: two of type SINALPHA, two of type ASYM+, two of type WMEAN and the remaining two of type QUANT.
Q5. What is the average utility of the three approaches?
SINALPHA got an average utility of 79.8%, outperforming ASYM+ (78.8%) and QUANT (62.3%), but underperforming the WMEAN approach, that achieved an average utility in the last 20 rounds of 83.3%.
Q6. How do the authors feel about the sigmoid curve?
the authors intuitively feel by graphically analysing the curve that it permits a probably too soft penalisation of partners that proved to be trustable but that failed the last n contracts.
Q7. What is the next phase of the research?
In fact, the next phase of their work would be dedicated to this topic, and to the inclusion of the erosion property of trust in their approach.
Q8. What was the first round of the selection of suppliers?
In the first 40 rounds, the selection of suppliers was done randomly, and in the last 60 rounds the selection was done taking into account the approach used by the buyers.
Q9. What are the two assumptions used in the sigmoid curve?
The authors use these two assumptions in their experiments, although their proposed aggregation engine might be extended in the future to more complex and diversified representation of trust information.
Q10. What is the main difference between the two types of trust dynamics?
In fact, in one experiment described in [19], a single negative observation that happens after a high number of previously observed positive experiences makes the trust level to decrease sharply, after which it takes a long sequence of positive observations to getting back to the previously trust value.
Q11. What was the interesting result of this experiment?
the most interesting result of this experiment was the capacity of SINALPHA buyers in adapting to the situation by massively choosing suppliers of type SB after round 70.
Q12. Why did the authors study the SINALPHA approach in population B?
In the experiments with population B, the authors intended to study the performance of the SINALPHA and the WMEAN strategies in the presence of extreme partners’ behaviour, particularly the cases where good partners, which have been successfully in fulfilling their obligations, suddenly start having systematic deceptive behaviour.
Q13. What is the main difference between the STexVM and the FIPA marketplace?
It follows the multi-agent paradigm, and is implemented over Jade platform, using the standard behaviours of Jade and FIPA performatives and interaction protocols4.