Trusted Dynamic Coalitions

Businesses are finding more and more ways to take advantage of the
internet, and can they can reduce costs and increase their sales by going
online. Communication across continents is easier than ever before, and so
they can also use the internet to work together. Smaller specialist
companies can now compete with large multinationals by
working together to distribute sub-tasks of the work to companies
specialist in that sub-task. They have then created a "virtual
organisation" -- an increasingly common mode of working.

Virtual organisations also form when organisations, military forces
and bystanders put aside differences and work together to alleviate
the consequences of a major disaster. One of the characteristics of
these virtual organisations is that they are highly fluid -- as the
situation on the ground changes, partners and the trust between them
changes, structures change and even the purpose of the virtual
organisation can change. In the military sphere, virtual
organisations such as these are known as dynamic coalitions.

A problem virtual organisations face is that the members often do not
trust one another. They may be competitors for the most part, and
only working together on a single job, and so there will naturally be
information that they are not prepared to disclose to potential
competitors.

In a virtual organisation, partners need to take action based on the
information they receive from other partners. This presents a problem,
because the information may not be trustworthy. To increase their
confidence that they are taking the right action, there are some
things a partner can do. They can ask the sender for more details
about the information itself, such as its source or age (known as the
provenance of the information), but these details can often give away
organisation or government secrets, and partners are naturally not
keen to release them. They could try to verify the information
themselves or via a third party, but these activities take time and
in situations such as disaster response, speed is an
important factor.

We wish to enable a partner to increase their confidence that they are
taking the right action in a virtual organisation. We will do this
by providing methods and tools that help partners to choose the
virtual organisation policies so that the desire for provenance data
and the desire for secrecy are held in balance.

We will look particularly at the provenance acquisition policy,
which states what provenance information is associated with
communications within the virtual organisation, and the provenance use
policy, which states the action that a partner will take on receipt of
information with a certain provenance associated to it.

Choosing the right combination of policies here is critical to the
success of a virtual organisation, and this choice must be made and
re-made as the partners, trust relationships and goals of the virtual
organisation change.

We will build mathematical models of the policies and the virtual
organisation itself, and a tool to allow partners to interact with
these models (without their needing to understand the models) to
predict the impact of changes to these policies in terms both of the
secrecy requirements within the virtual organisation and their ability
to have confidence in the information they receive. This will help in
the choice of provenance policies both at the start of a virtual
organisation and throughout its lifetime.

Direct impact will come from activities that will be carried out throughout the project. They include:

* Consistent engagement with Dstl. Our collaboration with Dstl will ensure that the project's outcomes will be tested on DC use cases that are of relevance to the security and defense community;

* Dissemination activities both at the academic level and beyond, including leveraging other EPSRC funded projects, to which the School of Computing is associated;

* Independent engagement of project members with standardization bodies within the provenance management community.

These activities are further detailed in the Pathway to Impact document, one section of which is devoted specifically to impact in the defence community.

One of the immediate goals of the project is to provide initial insight into the data privacy/transparency tension that characterises partner interactions in the context of Dynamic Coalitions where trust levels may vary, namely the need for information privacy on one side, against the need for rich metadata disclosure, on the other. Our goal is to provide the technical underpinning on which a framework can be built, by which partners can take account of the available trade-offs in this space, while negotiating partnership agreements. This has the potential to improve both the effectiveness and efficiency of the coalition, as well as the risk management capability of each of the partners with respect to data privacy.