Ontologies for Trust and Reputation
Ontology can be viewed as a shared conceptualisation of a domain that is commonly agreed to by all parties. In this research project, we define ontological manifestations of trust and reputation. We propose a generic ontology and a specific ontology for trust and reputation. We define trust ontology, agent trust ontology, service trust ontology and product trust ontology. The relationship between these ontologies is expressed. Additionally, we define reputation ontology, agent reputation ontology, service reputation ontology and product reputation ontology. The relationship between these ontologies is expressed. An ontological manifestation is defined for reputation relationship, reputation query relationship, recommender relationship and third party trust relation.
Fuzzy Intelligent System for Product and Service Design
Before introducing new products or new services into the market, product or service designers need to optimise marketing share in an eco-environment based on customers’ survey data. In traditional product and service design, various approaches have been proposed to achieve the goal based on analysis of customers’ survey data. Due to human perceptivity, fuzziness is always inevitable in customers’ survey data. However, the existing approaches are unable to deal with fuzziness in customers’ surveys, thus robustness of the product and service design cannot be addressed. The goal of this project is to enhance the existing approaches by considering the fuzziness in customers’ survey data such that the deficiency can be compensated. A fuzzy intelligent system for optimising customer satisfactions in eco-product & service design is proposed based on the three assemblies: (a) A fuzzy segmentation method for partitioning a large amount of customers into fuzzy clusters based on customers’ survey data; (b) A fuzzy modelling algorithm for relating the fuzzy customer satisfactions and the design attributes, which indicate the fulfilment of engineering characteristics; (c) A fuzzy multi-objective optimisation model for optimising market share and production cost constrained with engineering characteristics. The resulting fuzzy intelligent systems will be applied on eco-product & service design in which the customer satisfactions are intersected with ecological and economical values. A case study of freeway network monitoring systems is conducted to evaluate the effectiveness of the proposed fuzzy intelligent system on eco-product & service design.
Reputation-Based Decision Making
Reputation-based decision making is of particular importance for making decisions in virtual environments, especially when the trust information is not available. In this research project, we define the concept of reputation and reputation values, taking into account the context-dependent and time-dependent nature of reputation. Additionally, the concept of a reputation relationship is defined and expressed. We define mathematical expressions and frameworks that can be used to express the reputation value (repute) of an entity both qualitatively and quantitatively, both currently and in the future. Similar to its counterpart, complementary methodology, this methodology for calculating reputation value at the current time and at a point in time in the future takes into account various factors like the context, time, trend, seasonality and noise.
Maintaining Trust in Virtual Environments
The aim of this research is to propose and validate a methodology for maintaining trust in virtual environments. Trust is broadly acknowledged to be an important element for the efficient and effective operation of virtual business network. This is because trust functions like the glue that holds and links virtual network members together as they relate and collaborate remotely. Virtual business transactions are generally short lived, therefore trust needs to build quickly. However, this type of trust seems to be very fragile and dispersed easily. Despite the fragility of trust in virtual environments, it is challenging to work out how to maintain the level of trust (specifically positive trust) in any form of virtual relationship. Once positive trust has been established, it needs to be maintained at that level or should even be increased to a higher level. In order to achieve this objective, both the involved parties (trusting agent and trusted agent) need to make a coordinated and concerned effort. This thesis is an effort in that direction and proposes a methodology to achieve that goal. The proposed methodology would make use of CCCI metrics measure, iterative negotiation process and proactive performance evaluation measures such as control points. Validation of the proposed methodology will be carried out by computer simulation experiments using publicly available standard data sets. The outcome of this research can be used to direct or guide the interactions of virtual business entities to maintain their trust level in business relationships and it would help the sustaining of networked economy.
Trust-Based Decision Making
Trust-based decision making is of prime importance in virtual environments. Trust-based decision making is of particular importance when a given entity has a previous trust value regarding another entity. In this research project, we define the concepts of trust and trustworthiness considering the context-dependent and time-dependent nature of trust. Additionally, the concept of trust relationships and the association between the concepts of trust, trustworthiness and trust relationship are investigated. We define mathematical expressions and frameworks that can be used to express trustworthiness, both qualitatively and quantitatively at the current point in time and at a future time. The methodology for calculating trustworthiness value at the current time and in the future takes into account various factors like the context, time, trend, seasonality and noise.
Data Modelling for Public and Private Trust
In this project, a graphical modelling language is proposed for conceptual modelling of trust & trust relationship between agents in a service-oriented environment. The techniques are based on extension and semantic modification of various aspects of Unified Modelling Language (UML). This modified notation allows one to capture or express trust relationship modelling, trust property diagrams, trust context diagrams, trust transition diagrams, trust case diagrams, and trustworthiness assessment diagrams. The aim of the trust modelling language is the conceptual visualisation of a trust relationship that helps software engineers to communicate with business providers and end users & to automate the generation of a trust metric & a reputation metric for trust databases. This captures both static and dynamic aspects of trust & reputation.
Public Trust Infrastructure
Trust, reputation and risk data of an organisation can be used for the purposes of business intelligence primarily in the form of intelligent decision making. However, the trust, reputation and risk-related data are dispersed across various organisations and databases. The databases may be heterogeneous in nature. This project aims to develop a business intelligence infrastructure that would be used to store all the trust, reputation and risk information. The access to this information can be regulated by the person who owns the information. This would enable the sharing of business intelligence data (trust, reputation or risk) across organisations and across databases. An entity, instead of issuing reputation queries in order to make trust-based or reputation-based or risk based decisions, would proactively make use of the business intelligence infrastructure as a means of obtaining the data upon which it can base its decisions. This project additionally deals with the way in which the information stored in this public infrastructure can be made reliable for the decision making purposes.