An electronic voting system may be said to be composed by a number of components, each of which has a number of properties. One of the most attractive effects of this way of thinking is that each component may have an attached in-depth threat analysis and verification strategy.
Furthermore, the need to include the full system when making changes to a component is minimised and a model at this level can be turned into a lower-level implementation model where changes made can cascade to as few parts of the actual implementation as possible.
It appears that each researcher in the field of Electronic Voting Systems contributes to some particular aspect but rebuilds the whole system when they wish to implement this rather specific contribution.
The idea presented in this paper is that in order to build an e-voting system we simply add certain distinct pieces together – and in order to improve on a particular system we swap one distinct piece for another that fits into the same slot. In short, we are proposing that we start thinking about electronic voting systems as being component based.
Source: University of Surrey
Author: David Lundin