Information Technology Reference
In-Depth Information
liability specification (at contract negotiation time) to liability determination (at
litigation time). With the purpose of reduce legal uncertainty, the LISE approach
incorporates in B2B contracts (between services providers) an agreement about
the electronic evidence to be produced and used in case of failures. According
to this agreement, the parties commit to build these pieces of evidence, and
to rely on them for determining their share of responsibility. We assume that
certain conditions (such as, proof of authenticity and integrity) are satisfied by
complementary means.
In contrast with others frameworks that focus either in liability [17, 10] or
properties verification for logs [4, 11], we propose here an integrated frame-
work that allow us to specify liability, claims, and logs as electronic evidence
(Figure 1.a). As mentioned before we restrict ourselves in the context of liability
defined for a contractual environment. Furthermore, we exploit the central no-
tion of agents (seen as parties implied in a contractual engagement) that organize
claims, properties and distributed logs in a very tractable way. Then, we are able
to propose a general analysis procedure (Figure 1.b) allowing us to evaluate the
admissibility of a given claim instance. This procedure is focused on the notion
of properties attached to a given claim, rather than general properties describing
the behavior of the system [4, 11, 21, 3]. The fact that, in our context, claims are
defined a priori allow us to provide a simpler specification for such procedure.
We also study the incremental aspects of our procedure and propose alternative
solutions to obtain new results based on the results of a previous analysis.
(a)
(b)
Fig. 1. Framework for liability specification and analysis
This paper is dedicated to the presentation of our framework for the formal
representation and analysis of logs. We explore the aspects of the LISE methodol-
ogy introduced in [15] considering the distribution and analysis of logs. Another
previous work [16] studies how different distribution of logs can be classified with
relation to their content w.r.t. the level of interest that logging agents have in
changing the events in their logs. These can be viewed as complementary results
of the present work. Section 2 introduces a motivating example and some useful
notations. Section 3 presents our general model to specify logs, log distributions
and claims. Section 4 introduces some classical operations on distributed logs, as
extraction and merge. Finally, Section 5 gives the specification of our log anal-
ysis and its use for claim admissibility together with the studies of incremental
results.
 
Search WWH ::




Custom Search