Advisors: Dr. Wanita Sherchan (CSIRO ICT Centre, Canberra) / Dr. Surya Nepal (CSIRO ICT Centre, Sydney)
Bootstrapping Trust for Composite Services | |
| Description: |
A composite service is formed by combining existing services. A newly formed composite service does not have any trust value associated with it. Therefore, there is a need for mechanisms to bootstrap the trust for a composite service. A simple mechanism is to take an average or a mean of the trust values of the component services. However, this does not capture the complex relationships that exist among component services and factors influencing the trust of composite services. There has been a study on the bootstrapping of an atomic service using the concept of community. This may not be applicable to composite services as each of the component services may belong to different community. The project aims to achieve the following.
|
| Research Area: | Service-Oriented Computing, Reputation based Trust |
| Technologies: | Java, Eclipse, Web service technologies, etc. |
| Prerequisites: | Good Java programming skills are essential. |
Propagation of Trust in Supply-Chain like Composite Services | |
| Description: |
A composite service may build reputation over time by receiving feedback from its consumers. There are two types of composite services: vertically composed services and horizontally composed services (or supply-chain like composite services). We have developed an algorithm for propagating trust values in vertically composed services (propagation from a composite service to its component services). This algorithm may not be applicable to horizontally composed services. One of the issues that differentiates horizontal composition from vertical composition is the sequential dependencies of the component services. A sequential dependency is an influence of experience of one service on experience of subsequent services. This has a direct effect on the consumers’ perception of the performance of the composite service. The project aims to achieve the following.
It is important to note that we only consider reputation-based trust models in this project. |
| Research Area: | Service-Oriented Computing, Reputation based Trust, Sequential Effects, Service Composition |
| Technologies: | Java, Eclipse, Web service technologies, etc. |
| Prerequisites: | Good Java programming skills are essential. |
Implications of Service Composition on Reputation of Component Services | |
| Description: | A composite service is formed by combining existing component
services. The service composition has a huge implication on the
component services' trust since the trust value of a composite service
will be propagated to the component services. This project studies the
implications of service composition on reputation of each component
service. There have been several studies on implications of reputation
in partnerships in the fields of economics and business. This project
plans to borrow the ideas from the fields of business and economics and
apply to service composition. The project aims to achieve the following.
It is important to note that we only consider reputation-based trust models in this project. |
| Research Area: | Service-Oriented Computing, Reputation based Trust, Service Composition |
| Technologies: | Java, Eclipse, Web service technologies, etc |
| Prerequisites: | Good programming skills are essential. |
A Behaviour Based P2P Trust Model for Service Web | |
| Description: | Most of the reputation based trust models such as in e-commerce
marketplaces are largely seen as a way to penalise the service providers
and consumers who have been behaving badly. The aim of having trust
models is not only to penalise bad behaviour but also to reward and
encourage good behaviour to bring behavioural changes on the
participants towards improving their services. The latter part has been
largely ignored in the P2P trust models. This project aims to study the
behaviour of individuals in social field and develop a behaviour based
P2P trust model for the Service Web. The project aims to achieve the
following.
|
| Research Area: | Service-Oriented Computing, Reputation based Trust, Social behaviour analysis |
| Technologies: | Java, Eclipse, Web service technologies, etc. |
| Prerequisites: | Good programming skills are essential. |
An Efficient Query Processing Algorithm for the Semantic Service Web | |
| Description: | The queries on the Semantic Service Web have the following features:
(a) the environment is heterogeneous and distributed; (b) the query
results are fuzzy in nature; and (c) the query results are a combination
of ranked list of services. These features make the TA (Threshold
Algorithm)-style algorithms an ideal candidate for processing such
queries. The TA-style algorithms have been used widely to process top-k
queries in domains ranging from heterogeneous databases to information
retrieval. The aim of the project is as follows.
|
| Research Area: | Service-Oriented Computing, Query Processing, Semantic Service Web, TA-style algorithms |
| Technologies: | Java, Eclipse, Web service technologies, etc. |
| Prerequisites: | Good programming skills are essential. |
A Model for Provenance of Reputation in the Service Web | |
| Description: | Increasingly, provenance is considered an important
aspect of electronic data and is gaining prominent position in research. The users of
electronic systems would like to build confidence on the obtained
results. Provenance provides such confidence. This project aims to study
provenance of reputation in the Service Web. The project
aims to achieve the following.
|
| Research Area: | Service-Oriented Computing, Provenance, Reputation |
| Technologies: | Java, Eclipse, Web service technologies, etc. |
| Prerequisites: | Good programming skills are essential. |