=Paper=
{{Paper
|id=Vol-2938/paper-PROBLEMS-45
|storemode=property
|title=Process Mining with Common Sense
|pdfUrl=https://ceur-ws.org/Vol-2938/paper-PROBLEMS-45.pdf
|volume=Vol-2938
|authors=Diego Calvanese,Sanja Lukumbuzya,Marco Montali,Mantas Simkus
|dblpUrl=https://dblp.org/rec/conf/bpm/CalvaneseLMS21
}}
==Process Mining with Common Sense==
https://ceur-ws.org/Vol-2938/paper-PROBLEMS-45.pdf
Process Mining with Common Sense
Diego Calvanese1,2 , Sanja Lukumbuzya3 ,
Marco Montali1 , and Mantas Simkus3
1
Free University of Bozen-Bolzano, Bolzano (Italy)
{calvanese,montali}@inf.unibz.it
2
Umeå University, Umeå (Sweden)
3
Vienna University of Technology, Vienna (Austria)
sanja.pavlovic@tuwien.ac.at, simkus@dbai.tuwien.ac.at
Abstract. We argue that, with the growth of process mining in breadth
(variety of covered tasks) and depth (sophistication of the considered pro-
cess models), event logs need to be augmented by commonsense knowl-
edge to provide a better input for process mining algorithms. This is
crucial to infer key facts that are not explicitly recorded in the logs, but
are necessary in a variety of tasks, such as understanding the event data,
assessing their compliance and quality, identifying outliers and clusters,
computing statistics, and discovering decisions, ultimately empowering
process mining as a whole.
Keywords: Commonsense knowledge · process mining · event data
1 Description of the problem
Process mining is experiencing a golden age, with a flourishing academic commu-
nity and an increasing adoption by industry. The field evolved from the seminal
works on control-flow process discovery from the late 90s to a full-fledged arsenal
of techniques for process improvement. Contemporary process mining techniques
indeed range from advanced forms of process discovery to sophisticated forms of
process enhancement, conformance checking, and online operational support. At
the same time, such techniques have been refined and extended over the years to
target increasingly sophisticated process models. At the frontier of research, we
indeed find models that tackle multiple perspective at once (such as control-flow,
data, and decisions - see, e.g., [16, 5, 17, 10, 13]), and that are moving their focus
from single cases to multiple, co-evolving objects [1, 12, 4, 18, 14].
Mirroring this trend, the event data constituting the input fuel for process
mining algorithms evolved from simple (multi)sets of traces containing sequences
of propositional events to richer traces annotated with different types of data
attributes [16, 6, 15], finally culminating in full-fledged networks of events related
to each other through multiple objects and relationships [2, 11]. This evolution is
also witnessed by the employed data formats from event logs, moving from the
well-established XES (https://xes-standard.org) standard to recent propos-
als such as XOC [15] and OCEL (http://ocel-standard.org).
Copyright © 2021 for this paper by its authors. Use permitted under
Creative Commons License Attribution 4.0 International (CC BY 4.0).
2 D. Calvanese et al.
In this complex spectrum, it is common practice to make the assumption
that the event log used as input for a process mining project explicitly contains
enough relevant facts to faithfully apply the intended algorithms. We argue that,
even when the recorded event data are of high quality, this is in general a too
strong assumption. Going a step further, a human, or, to be more precise, a
domain expert, would be able to easily reconstruct all the relevant facts from the
log by applying their own commonsense knowledge.4 The following two examples
ground this observation in two relevant settings: reconstruction of the state of
affairs in a process execution, and identification of outlier vs impossible traces.
Example 1. We focus on a typical object-centric process dealing with an order-
to-cash scenario where multiple orders may be handled together, possibly trans-
ferring order lines from one to the other, and applying coupons for getting a
discount. We assume that there is a database listing the different available item
types, and indicating, for each item type, the corresponding unit cost. Consider,
in this context, the following sequence of events (working on two, related orders,
where we omit the responsible persons for simplicity)5 :
event timestamp order item type attrs
create order 01/06/2021 9:00 o1
add item 01/06/2021 9:01 o1 laptop
create order 01/06/2021 9:03 o2
add item 01/06/2021 9:04 o1 mouse
add item 01/06/2021 9:08 o2 laptop
transfer content 01/06/2021 9:15 o2 , o 1
insert coupon 01/06/2021 9:20 o1 20%
insert coupon 01/06/2021 9:21 o1 20%
pay 01/06/2021 9:25 o1
By looking into this log and the price table mentioned above, a domain ex-
pert may easily reconstruct the final state of each order, including the overall
amount associated to order o1 that is paid when executing the last entry. This
is, in turn, relevant for a number of process mining tasks, such as classification
of orders and clustering of their corresponding traces, or mining decisions to un-
derstand which rules are applied by customers to choose under which conditions
an order is eventually paid or not.
In the specific sample log, what happens is that two orders are created con-
currently, only later on realising that they can be merged into a single order,
which is then finally paid. Specifically, the content of order o2 is transferred into
order o1 , which ultimately contains two laptops and one mouse, and is paid after
applying a discount.
4
We use “commmonsense knowledge” as an umbrella term for general knowledge
about the world, as well as general knowledge about a specific organisational setting.
5
We use a tabular format that resembles the input format used in [2] for discovery.
Process Mining with Common Sense 3
How is it possible to reconstruct these important, implicit facts, which are
not explicitly contained in the (sound and complete) given sequence of events?
Thanks to the application of commonsense knowledge (both general and domain-
specific). In fact, commonsense knowledge is used to understand that:
• when adding an item to an order, the other, previously added items will con-
tinue to be included in that order;
• when the content of order o2 gets transferred into another order o1 , the effect
is that o2 becomes empty, while o1 contains all and only those items that were
contained in o1 or o2 .
While these two examples relate to general, commonsense knowledge, there
is also domain-specific commonsense knowledge that only domain experts have.
Two key pieces of domain-specific knowledge in our example would be needed
to answer the following questions:
• What is the resulting state of an order after its content is being transferred
to another order? Does it stay active as an empty order? Or is it implicitly
cancelled?
• Is it possible to apply multiple discount coupons to the same order? If so, are
the discount percentages applied in cascade, or always to the overall amount
of the order?
Notice that reconstructing the state of affairs and answering the questions
above are orthogonal aspects to the quality of event data, as they concern implicit
facts related to the cumulative effect of events, not the explicit events themselves.
The next example is inspired by [3], though there they focus on understanding
process modeling constructs, whereas here we keep our focus on event data.
Example 2. Consider a simplified version of a case-centric logistic process where
each process instance focuses on the prepare- of a single package. We do not
have data attributes, just sequences of events (with their frequency). Consider
in particular the following log, containing three process variants:
Variant 1 (970 traces) Variant 2 (20 traces) Variant 3 (10 traces)
receive payment
prepare package prepare package prepare package
load package load package
deliver package deliver package deliver package
receive payment receive payment
The most common Variant 1 represents the standard execution of the
process: after receiving the payment from the customer, the package is pre-
pared, loaded in a truck, and delivered to the customer. Variant 2 represents
an outlier behaviour, where a package is prepared and delivered without a
prior payment, which is in fact received only after the delivery is completed.
Interestingly, a domain expert could judge that this behaviour is an outlier
behaviour event without looking at the frequency, if it is common practice in
the company to only deliver material that has been paid before. This shows the
multitude of usages of commonsense knowledge.
4 D. Calvanese et al.
What discussed for Variant 2 is taken to the extreme when looking at Vari-
ant 3. There, a package is prepared without a prior payment. What appears even
stranger, though, is that the package is not loaded in the truck, but is never-
theless delivered, finally resulting in a payment. This appears strange because,
again due to commonsense knowledge, a human would immediately understand
that a package cannot be delivered in a truck if it has not been loaded there up-
front. This, in turn, would immediately lead to label Variant 3 as a variant that
cannot be simply judged as an outlier, but requires instead further inspection to
understand whether it relates to: incomplete/faulty logging of activities (e.g., the
package was actually loaded, but this operation has not been logged), presence
of a fraud (money received without any material delivery), or other. Under-
standing what is happening with Variant 3 is in turn crucial towards compliance
assessment, classification of outliers, computation of statistics, data quality, etc.
The main issue is that this commonsense knowledge is not explicitly com-
municated to process mining algorithms: it either stays in the head of domain
experts, and only indirectly emerges in how such experts interact with such al-
gorithms and the results they produce, or it gets embedded into ad-hoc, hardly
interpretable pieces of code used to (pre-)process event data. At the same time,
eliciting such knowledge is a notoriously difficult problem.
All in all, the problem is then: How can we suitably augment event data
with commonsense knowledge, to improve the faithfulness and quality of process
mining without introducing too much additional modelling effort?
2 Why is this an important problem
This problem relates to the grand challenge of garbage-in garbage-out in pro-
cess mining and, more in general, in data science. While the community has
extensively worked on data quality targeting the data explicitly contained in an
event log, no attention has been devoted to the insights that can be obtained
through commonsense reasoning on such data. The two examples above indicate
that even in very simple examples this is of utmost importance to support, and
better inform, process mining algorithms.
At the same time, the elicitation and usage of commonsense knowledge is
one of the central open problems in (general) artificial intelligence [8]. Focusing
on a concrete setting, such as that of process mining, grounds this problem in
a concrete context, paving the way to more accessible results that could in turn
provide insights on how to advance with the problem in its full generality.
3 Relation with existing work
Within the BPM community, this problem is largely unexplored. The literature
has so far mainly targeted the problem of integrating structural knowledge with
process models (see, e.g., [4]), or in providing richer ontological foundations for
Process Mining with Common Sense 5
process models themselves (see, e.g., [3]). The huge body of work on data qual-
ity for event data is complementary, and in synergy, with the problem described
here. Progress within artificial intelligence on commonsense knowledge and com-
monsense reasoning, in terms of general inference mechanisms [7] and knowledge
bases targeting specific domains that are relevant for BPM [9], provides a solid
basis to attack the problem presented here.
4 Initial ideas towards solving the problem
As pointed out in [8], commonsense reasoning is usually tackled using either
knowledge-based or machine learning-based approaches, with very limited in-
teractions between them. Also crowd-sourcing approaches exist. In the setting
considered here, we advocate the knowledge-based approach, which is particu-
larly effective when dealing with qualitative reasoning [8], of central importance
when applying common sense to event data.
Specifically, we are investigating the adoption and further development of
techniques in the area of Knowledge Representation & Reasoning (KR&R).
KR&R is developing methods for capturing human knowledge in machine-
processable knowledge bases, which can used by automated reasoning engines to
provide non-trivial insights to the users. If knowledge bases consist of raw data
enriched with common sense knowledge, then reasoning engines can be used
to infer useful new facts that follow logically from the data and the captured
human knowledge. This can for example be exploited to alleviate information
incompleteness and to identify inconsistencies in the data. KR&R already offers
several techniques related to the challenges discussed above, such as reasoning
about actions and change, non-monotonic reasoning, rule-based languages, and
description logics. These individual techniques have different strengths, which
need to be combined in order to address the multifaceted needs discussed above.
As a first step, we are developing a new lightweight formalism that combines
structural knowledge (to represent the relevant classes and relationships within
an organisation), temporal/dynamic operators (to represent the process dynam-
ics), and nonmonotonic features (essential to capture what does not change when
an event occurs), while guaranteeing efficient inference mechanisms. To the best
of our knowledge, a formalism balancing all these different ingredients is missing.
The second step will be to understand how minimal knowledge bases ex-
pressed in this formalism can be elicited and (re)used to augment event data,
analysing the impact on applicability and quality of process mining techniques.
Acknowledgements. This research has been partially supportedby the Wal-
lenberg AI, Autonomous Systems and Software Program (WASP) funded by the
Knut and Alice Wallenberg Foundation, by the Italian Basic Research (PRIN)
project HOPE, by the EU H2020 project INODE, grant agreement 863410, by
the CHIST-ERA project PACMEL, by the project IDEE (FESR1133) funded
by the European Regional Development Fund (ERDF) Investment for Growth
and Jobs Programme 2014-2020, and by the Free University of Bozen-Bolzano
through the projects KGID, GeoVKG, STyLoLa, and VERBA.
6 D. Calvanese et al.
References
1. van der Aalst, W.M.P.: Object-centric process mining: Dealing with divergence
and convergence in event data. In: Proc. of SEFM. LNCS, vol. 11724, pp. 3–25.
Springer (2019)
2. van der Aalst, W.M.P., Berti, A.: Discovering object-centric Petri Nets. Funda-
menta Informaticae 175(1-4), 1–40 (2020)
3. Adamo, G., Di Francescomarino, C., Ghidini, C., Maggi, F.M.: Beyond arrows
in process models: A user study on activity dependences and their rationales.
Information Systems 100, 101762 (2021)
4. Artale, A., Kovtunova, A., Montali, M., van der Aalst, W.M.P.: Modeling and rea-
soning over declarative data-aware processes with object-centric behavioral con-
straints. In: Proc. of BPM. LNCS, vol. 11675, pp. 139–156. Springer (2019)
5. Burattin, A., Maggi, F.M., Sperduti, A.: Conformance checking based on multi-
perspective declarative process models. Expert Systems with Applications 65, 194–
211 (2016)
6. Calvanese, D., Kalayci, T.E., Montali, M., Tinella, S.: Ontology-based data access
for extracting event logs from legacy data: The onprom tool and methodology. In:
Proc. of BIS. LNBIP, vol. 288, pp. 220–236. Springer (2017)
7. Davis, E.: Logical formalizations of commonsense reasoning: A survey. J. of Arti-
ficial Intelligence Research 59, 651–723 (2017)
8. Davis, E., Marcus, G.: Commonsense reasoning and commonsense knowledge in
artificial intelligence. Communications of the ACM 58(9), 92–103 (2015)
9. Davis, E., Marcus, G., Frazier-Logue, N.: Commonsense reasoning about containers
using radically incomplete information. Artificial Intelligence 248, 46–84 (2017)
10. De Smedt, J., Hasic, F., vanden Broucke, S.K.L.M., Vanthienen, J.: Holistic dis-
covery of decision models from process execution data. Knowledge Based Systems
183 (2019)
11. Esser, S., Fahland, D.: Multi-dimensional event data in graph databases. CoRR
Technical Report arXiv:2005.14552, arXiv.org e-Print archive (2020), https://
arxiv.org/abs/2005.14552
12. Fahland, D.: Describing behavior of processes with many-to-many interactions. In:
Proc. of Petri Nets. LNCS, vol. 11522, pp. 3–24. Springer (2019)
13. Felli, P., Gianola, A., Montali, M., Rivkin, A., Winkler, S.: CoCoMoT: Confor-
mance checking of multi-perspective processes via SMT. In: Proc. of BPM. LNCS,
vol. 11157, pp. 219–235. Springer (2021)
14. Ghilardi, S., Gianola, A., Montali, M., Rivkin, A.: Petri nets with parameterised
data - modelling and verification. In: Proc. of BPM. LNCS, vol. 12168, pp. 55–74.
Springer (2020)
15. Li, G., de Murillas, E.G.L., Medeiros de Carvalho, R., van der Aalst, W.M.P.:
Extracting object-centric event logs to support process mining on databases. In:
Proc. of CAiSE Forum. LNBIP, vol. 317, pp. 182–199. Springer (2018)
16. Lu, X., Nagelkerke, M., van de Wiel, D., Fahland, D.: Discovering interacting
artifacts from ERP systems. IEEE Trans. Serv. Comput. 8(6), 861–873 (2015)
17. Mannhardt, F., de Leoni, M., Reijers, H.A., van der Aalst, W.M.P.: Data-driven
process discovery - revealing conditional infrequent behavior from event logs. In:
Proc. of CAiSE. LNCS, vol. 10253, pp. 545–560. Springer (2017)
18. Polyvyanyy, A., van der Werf, J.M.E.M., Overbeek, S., Brouwers, R.: Information
systems modeling: Language, verification, and tool support. In: Proc. of CAiSE.
LNCS, vol. 11483, pp. 194–212. Springer (2019)