Quickening Data-Aware Conformance Checking through Temporal Algebras

Abstract

A temporal model describes processes as a sequence of observable events characterised by distinguishable actions in time. Conformance checking allows these models to determine whether any sequence of temporally ordered and fully-observable events complies with their prescriptions. The latter aspect leads to Explainable and Trustworthy AI, as we can immediately assess the flaws in the recorded behaviours while suggesting any possible way to amend the wrongdoings. Recent findings on conformance checking and temporal learning lead to an interest in temporal models beyond the usual business process management community, thus including other domain areas such as Cyber Security, Industry 4.0, and e-Health. As current technologies for accessing this are purely formal and not ready for the real world returning large data volumes, the need to improve existing conformance checking and temporal model mining algorithms to make Explainable and Trustworthy AI more efficient and competitive is increasingly pressing. To effectively meet such demands, this paper offers KnoBAB, a novel business process management system for efficient Conformance Checking computations performed on top of a customised relational model. This architecture was implemented from scratch after following common practices in the design of relational database management systems. After defining our proposed temporal algebra for temporal queries (xtLTLf), we show that this can express existing temporal languages over finite and non-empty traces such as LTLf. This paper also proposes a parallelisation strategy for such queries, thus reducing conformance checking into an embarrassingly parallel problem leading to super-linear speed up. This paper also presents how a single xtLTLf operator (or even entire sub-expressions) might be efficiently implemented via different algorithms, thus paving the way to future algorithmic improvements. Finally, our benchmarks highlight that our proposed implementation of xtLTLf (KnoBAB) outperforms state-of-the-art conformance checking software running on LTLf logic.

Streamlining Temporal Formal Verification over Columnar Databases

“…Please consider that Declare clauses do not necessarily reflect association rules, as the latter do not provide temporal constraints correlating the activation of activation and target conditions. In this paper, we focus on Declare clauses only predicating over the events' activity labels, which are then referred to as dataless; on the other hand, dataful Declare clauses can express data payload conditions over both activation and target conditions, as well as representing Θ payload correlation conditions between activating and targeted conditions [28]. Thus, both clauses and logs are referred to dataful otherwise.…”

“…KnoBAB [28,34] is a column database store tailored for both loading dataful logs being represented in XES [38] and dataless ones described as a tab-separated file. This outper-formed the previous state of the art in terms of both specification mining [39] and formal verification [35] tasks on tailored non-database solutions.…”

“…This was sensibly different from competing approaches [40,41] also relying on main memory engines where, instead, the query plan associated to a formal verification task is always expressed in terms of trees, thus not allowing the detection of shared sub-expressions to be merged to avoid wasteful recomputations. As vertices for a DAG can be sorted topologically, we can obtain for free the scheduling order in which the operators must be executed and, by associating each node a maximum distance value from the root, we can safely run in parallel all the operators laying at the same depth level, as all the previously called operators will pertain their information in an intermediate cache, thus achieving intraquery parallelism as a free meal [28]. This parallelization approach greatly differs from straightforward parallelization algorithms known in the Business Process Management area, where they simply run each declarative clause occurring in the specification in a separate thread [35].…”

“…In the context of formal specification tasks expressed in LTL f , recent research clearly remarked on the inadequacy of off-the-shelf row-based relational databases and SQL as a query language for expressing LTL f temporal constraints, as it clearly showed that a customized relational algebra for expressing formal specification (eXTended LTL f , xtLTL f [28]) and query plan minimizing the running of sub-queries [29] running on customized columnbased storage (KnoBAB [28,30]) outperformed the previous solution. The main benefit of this approach is that any LTL f can be directly expressed in terms of xtLTL f , while highlevel and human-readable temporal constraints expressed through temporal clauses can be directly specified in a semantics query at warm-up, thus allowing the support of any declarative temporal language (queryplan in Figure 1).…”

“…Concerning our datasets of interest, we only consider ones where trace events are not associated with a data payload, and therefore even such logs can be considered as dataless. On the other hand, with reference to Equation (1), event payloads can store patient information, thus registering the recorded medical condition being observed [28]; in the context of good brokerage, such payload might contain the relevant contract information between the supplier and the customer which are required to be respected (e.g., delivery times), as well as the location of the goods, their number, and quality [11].…”