An Approach to Developing Ontology-Based Tools
for Event Series Analysis
Anton Platunov
1a
, Lyudmila Lyadova
1b
, Nada Matta
2c
, Viacheslav Lanin
3d
and
Elena Zamyatina
1e
1
Department of Information Technologies in Business, HSE University, Perm, Russian Federation
2
University of Technology of Troyes, Troyes, France
3
Perm State University, Perm, Russian Federation
Keywords: Event Series, Data Analysis, Event Attributes, Event Logs, Process Mining, Multifaceted Ontology,
Rules Design, Rules Interpreter, Event Log Generation.
Abstract: Existing process mining methods allow to investigate processes in different domains. Besides mandatory
event attributes like as identifier, activity, and timestamp, additional event attributes can be present in data
sources. The analysing dynamics of changing the values of additional attributes allows to get important
information on the system. The applications must be developed by programmers with programming languages
to implement new methods of analysis. An approach to develop tools based on the use of algorithm designers
and expression builders like those included in MS Office applications is proposed in the paper. Their use does
not require programming skills. The implementation of the approach is based on a multifaceted ontology,
including descriptions of the rules for developing functions, as well as a description of functions for generating
and analysing event logs in accordance with these rules. The user interface for developing rules and the
algorithm for their interpreting are implemented in the research prototype of the application.
1 INTRODUCTION
Typically, Process Mining methods are used to
analyse the business processes of enterprises, where
it is possible to obtain structured information from
user workplaces to generate event logs. However, at
present, the scope of these methods has expanded
significantly: they use to solve the problems of
analysing social networks, analyse processes in the
healthcare process organization, etc. New
applications address the challenges of extracting data
from unstructured or semi structured heterogeneous
information sources to generate event logs. Another
feature of these domains is that additional attributes
specific to the appropriate domains are defined for
events. Analysing the values of these event attributes
allows to investigate the dynamics of the behaviour
of complex systems and to evaluate developing a set
of process in them.
a
https://orcid.org/0009-0008-7500-6278
b
https://orcid.org/0000-0001-5643-747X
c
https://orcid.org/0000-0001-8729-3624
d
https://orcid.org/0000-0002-0650-2314
e
https://orcid.org/0000-0001-8123-5984
New analysis methods are implemented through
the development of programs, in particular plugins
for the ProM system. Researchers must be proficient
in the programming languages (Java, Python, etc.)
which used for the applications developing, or must
involve professional programmers, specialists in
information technologies in the methods designing.
This approach to application development also
complicates the implementation of methods, slows
down research.
The task of developing tools that implement the
low-code principle becomes relevant. An approach to
creating a knowledge-driven analytical platform
based on domain specific modelling (DSM) is
proposed. Visual domain specific languages (DSL) to
develop algorithms and research scenarios can be
used. These languages correspond domain
specificity, reflect object characteristics and domain
limitations. Language toolkits are included in the
Platunov, A., Lyadova, L., Matta, N., Lanin, V. and Zamyatina, E.
An Approach to Developing Ontology-Based Tools for Event Series Analysis.
DOI: 10.5220/0012259100003598
In Proceedings of the 15th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2023) - Volume 2: KEOD, pages 323-330
ISBN: 978-989-758-671-2; ISSN: 2184-3228
Copyright © 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)
323

platform software. The using DSLs is available to
researchers, specialists in specific subject areas who
do not know professional software development
tools. The advantages of this approach are flexibility,
the ability to dynamically configure to solving new
tasks and to changing requirements. The problem of
creating such language toolkits is the need to
implement model editors and model transformation
tools or model interpreters.
An alternative approach is to develop tools based
on the use of algorithm designers and expression
builders like those included in MS Office
applications. These tools are available for users who
have the skills to work with the appropriate MS
Office tools, where users can develop own complex
algorithms for processing and analysing data using
standard application functions, and the existing set of
functions can be expanded by users (like the plugins
development).
The challenge of the project is creating a research
prototype of tools based on the low-code principles
for analysing event series with attributes of various
types.
The tasks of the approach implementation:
1. Developing a language to describe user
functions and research scenarios.
2. Designing representation of functions and
research scenarios, defining data structures for
implementing them.
3. Developing a parsing and interpretation
algorithms for the functions and research
scenarios.
4. Designing a user interface for function builder.
This paper focuses on the solving these tasks.
2 RELATED WORKS
Let's review approaches to solving the most
interesting tasks in development of process analysis
tools considering the values of event attributes, the
dynamics and behaviour of these attributes.
The papers (Cremerius, 2021; Cremerius, 2022;
Cremerius, 2023) consider the issues of analysing
processes with methods taking into account the values
of attributes specific to domain and their behaviour.
The authors note that besides mandatory event
attributes like as identifier, activity, and timestamp,
additional event attributes can be present in data
sources. For example, researchers can analyse
additional attributes of events specific for their
domain such as human resources, costs, and
laboratory values. The analysis of attribute values is
especially important in healthcare processes, where a
huge amount of data is generated (such as vital signs
or laboratory measurements representing a patient’s
wellbeing). Analysis of developing these
measurements depending on which treatment
activities were conducted can help to evaluate
different treatment paths, to select one that might
result in better patient wellbeing than others. Process
mining traditionally focusses on the workflow control
aspects. However, processes not only include
activities and their orderings, but also the data
generated and manipulated during process
executions. Any process activity generates data, but
this information does not play the role in process
mining that it deserves. Data-enhanced process
models are described by authors. They give formal
definitions of event, event log, trace, process model
and process variants. Event attribute selection and
aggregation are defined. Authors focus on healthcare
processes. Nevertheless, the developed methods can
be applied to different domains, where event
attributes are available.
In the papers (Bano, 2020; Bano, 2021) authors
also argues that despite event logs capturing
behavioural information (most discovery algorithms
focus on process control-flow), they are a rich source
of domain specific data useful for analysis data-flow
perspective. Usually, this data is not represented
explicitly in process models, but it provides valuable
contextual information. A semi-automatic approach
to discover a data model that complements traditional
process mining techniques with domain specific
information is proposed.
Another important problem is the processing of
uncertainty of the event attribute values
(Pegoraro, 2019). In the paper the setting of uncertain
event logs is analysed: quantified uncertainty in the
logs is recorded together with the corresponding data.
Authors define a taxonomy of uncertain event logs
and models and examine the challenges that
uncertainty poses on process discovery and
conformance checking. They show how upper and
lower bounds for conformance can be obtained
aligning an uncertain trace onto a regular process
model.
The paper (Kampik, 2022) presents the results of
survey of challenges and perspectives around event
log generation in process mining. The responses of an
industry experts analysed by authors indicate that
particularly relevant challenges exist in data
integration and quality. Authors argue that process
mining can benefit from a systematic integration
traditional methods and wide-spread business
intelligence approaches.
The problems of generating event logs
KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development
324

(Mitsyuk, 2017) and their preprocessing (Marin-
Castro, 2021) are devoted to the articles that consider
tasks of forming event logs based on data obtained
from various sources (from data bases
(Calvanese, 2016) and network messages of trading
systems (Carrasquel, 2021), social networks
(Lanin, 2021; Peña-Araya, 2015; Ritter, 2012) and
mass media (Abrosimova, 2018; Shalyaeva, 2016;
Shalyaeva, 2017), etc.).
However, the implementation of the proposed
methods requires the involvement of professional
developers who have skills of using software
development tools and programming languages
(Java, Python, etc.). It complicates development and
slows down research with using Process Mining
tools.
The authors of papers (Lyadova, 2022; Zayakin,
2023) propose an approach to developing an
ontology-driven analytical platform that includes
toolkits for creating visual domain specific languages
(DSL) that are customised to solve tasks in research
domains. Created languages reduce the requirements
for knowledge and skills for researchers in the field
of software development. The use of DSL minimizes
errors because domain restrictions are included in
DSL, language semantics complies with
characteristics of domain objects and corresponds to
existing domain restrictions.
An approach to solving the problem of the
creating DSL through automating DSL metamodels
generation, customising languages based on the
multifaceted ontology, is proposed in the paper
(Kulagin, 2022).
However, this approach also requires developers
to have skills in knowledge engineering, in formal
languages and grammars that are necessary for
creating metamodels of basic languages, developing
domain ontologies. An alternative approach to
developing ontology-driven tools for generating
event logs and analysing event series that do not
require programming skills is proposed in this paper.
3 ONTOLOGY-BASED TOOLS
FOR THE EVENT SERIES
ANALYSIS
The requirements for the developed tools are
determined based on the descriptions given in the
paper (Zayakin, 2022), which provides event-time
series and event series definitions. Examples of
generating event logs based on data, extracted from
various sources, according to rules described by
users, as well as examples of rules for analysing
processes, based on event logs containing additional
attributes, are included in paper.
The designed tools shall meet the following
functional requirements:
1. Determining the data sources that will be used
to create event series containing additional attributes
(numerical indicators).
2. Describing the rules for generating event logs
based on numerical attributes processing.
3. Describing the rules for calculating numerical
event characteristics based on data, extracted from
various sources, and parameters, set by users, and
aggregating these characteristics with events.
4. Interpreting rules defined by users.
The event log structure must meet the XES format
used in ProM. Rules described by users should be
stored in the ontology.
3.1 Description of Functions for
Generation and Analysis of Event
Series
To implement functions for generation and analysis
of event series, a formal language for describing rules
should be developed.
The description of the rules (functions for
generating and analysing events) is based on the
definition of the superposition of functions.
The algorithm for solving any task can be
represented as a function that transforms the input
data into a result. The results of function calculation
can be used as input data for calculating other
functions. Thus, any complex algorithms can be
implemented as superposition of functions. It is
needed to define set of basic functions to implement
this approach.
This definition allows to simplify the grammar of
the language for describing rules and to develop a
universal user interface for developing and
interpreting rules.
When developing a language, two types of rules
should be distinguished, which may have structural
differences, therefore, each of them should be
described separately:
1. Rules for calculating events for generating
event logs. The user defines rules by which the types
of events are determined based on the values of input
data (Table 1). The result of applying these rules is
event log containing events with calculated event
types and timestamps (Table 2).
2. Rules for processing event logs extended with
numeric attributes. The user defines rules that allow
analysing behaviour of attributes associated with
An Approach to Developing Ontology-Based Tools for Event Series Analysis
325

events in the event log. The results of the analysis
according to such rules are included in the process
model (calculated characteristics are associated with
events).
Table 1: Example of input data (COVID-19 incidence).
Table 2: Examples of events determined based on incidence
rates.
The requirements for the language intended for
description of rules for generating and analysing
event logs are determined by the requirements for the
user interface, which should ensure the simplicity,
availability of the developed tools for non-
programmer users: rules should be developed in
designers that have a unified interface close to the
interface of designers (expression builders) in
Microsoft Office applications.
The rule definitions can include built-in
operations and functions implemented in the
ontology. The operands of these function can be
numbers and text strings, arrays of values, time series,
event logs, the structures of which are defined in the
ontology, as well as the results of calculating
functions defined in the system.
Thus, rules (functions) described by users to
generate event logs using data obtained from
specified sources and to preprocess event logs with
additional attributes are defined as a superposition of
functions:
f = σ(f
0
, f
1
, ..., f
n
),
where the function f is defined as a superposition of
the functions f
0
, f
1
, ..., f
n
; and functions f
1
, ..., f
n
have
their own parameter sets for calculation:
f
1
(x
1
1
, x
1
2
, ..., x
1
k 1
), …, f
n
(x
n
1
, x
n
2
, ..., x
n
k
n
), and the
result is calculated as a function
f (x
1
1
, x
1
2
, ..., x
1
k 1
, …, x
n
1
, x
n
2
, ..., x
n
k
n
) =
= f
0
(f
1
(x
1
1
, x
1
2
, ..., x
1
k 1
), …, f
n
(x
n
1
, x
n
2
, ..., x
n
k
n
)).
All the functions are partial, that is, not
everywhere defined – there can be such combinations
of argument values for which the values of functions
f (x
1
1
, x
1
2
, ..., x
1
k 1
, …, x
n
1
, x
n
2
, ..., x
n
k
n
) do not exist: at
least one of the values f
1
(x
1
1
, x
1
2
, ..., x
1
k 1
), …,
f
n
(x
n
1
, x
n
2
, ..., x
n
k
n
) does not exist; or these values exist
and are equal to b
1
, ..., b
n
, but does not exist value of
f
0
(b
1
,..., b
n
).
Thus, the method of describing the algorithm for
calculating a function, the type of rules is defined.
To describe the rules for generating event logs and
their preprocessing, a language has been developed.
The formal language grammar includes more than
50 syntactic rules that define non-terminal symbols of
the language. The language defines, along with
standard types for programming languages, data types
for describing time stamps, time series, events with
additional attributes and event logs that can be
processed or generated using functions, etc. The
target grammar symbol is “Function”.
The designed functions include operations on
values of standard types (arithmetic operations,
operations on text strings, comparison operations,
etc.) and built-in functions (MAX, MIN, AVG, SUM,
etc.). To simplify the interface of the expression
builder, operations AND, OR, XOR are implemented
as functions by analogy with the corresponding
functions of MS Excel (this simplifies the considering
operation priorities when forming expressions and
parsing the user-defined function, when interpreting
(calculating) the built function). The expression
designer also allows to use the functions SWITCH or
IFTHEN (to determine alternative calculations
depending on the specified conditions), FOREACH
(to process data in a loop).
The grammar described is of type LL(1).
Based on the described rules, algorithms for
parsing and interpreting functions described by users
have been developed. When analysing the
descriptions of functions, the algorithm of left-
recursive descent is implemented.
KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development
326

The grammar of the rule description language is
developed using Backus-Naur forms. However, a
representation in the form of Wirth diagrams is more
suitable for including rules in the ontology.
3.2 Ontology of Rules (Functions)
The developed grammar of the language for function
description (rules for generating and analysing event
logs) includes more than fifty syntactic rules.
Grammar rules are included in the ontology.
Ontology of rules contains descriptions of two
type functions (rules defined by users). Fragment of
the main classes is shown in Figure 1 and Figure 2.
An example of a rule description for calculating
event types when generating an event log is shown in
Figure 3.
When analysing the descriptions of functions
developed by users according to the rules included in
the ontology, a representation of algorithms for
calculating functions in the form of a tree is built
(Figure 4). The interpretation (function calculation) is
implemented through traversing the constructed tree.
Figure 1: Fragment of the class hierarchy of rules.
Figure 2: Classification of rules as ontograph fragment.
Figure 3: An example of a rule description as ontograph.
Figure 4: An example of the algorithm for calculating function XOR represented as a tree.
An Approach to Developing Ontology-Based Tools for Event Series Analysis
327

3.3 Implementing Research Prototype
The research prototype is developed in C#. A user
interface is implemented to develop rules (functions)
for generating and analysing event logs.
The research prototype structure is shown in
Figure 5. The modules of the application implement
the functions described above.
The user form of the event calculation rule builder
is presented in Figure 6. Using the drop-down list, the
user determines the rule type (event calculation rule
or event log processing rule). User describes the list
of formal parameters where name for each parameter
and type are defined (for calculating events it is
TimeSeries type; for the event log it is EventLog
type). User can add formal time series constraint
parameters to the timestamp using the Add
Boundaries buttons. Also, the user can write a
comment that will describe the rule. Alternative
calculation options can be specified for different
conditions. The form controls allow to define all
elements of functions defined in grammar rules.
Figure 5: Structure of the research prototype.
Figure 6: The user form of the event calculation rule builder.
KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development
328

Figure 7: The result of generating an event log.
The result of generating an event log according to
the rules specified by user is presented in Figure 7.
The results of generating and processing event
logs according to the rules specified by users can be
uploaded to files in XES format (Figure 8) and
exported for analysis to ProM.
Figure 8: The result of generating an event log (fragment).
4 CONCLUSIONS
The main result is a research prototype of tools for
generating and analysing event logs. Experiments
have shown the practical significance of the proposed
approach. The developed tools have demonstrated the
universality of the created formal language,
flexibility, and accessibility of the software for users.
The developed tools can be integrated with process
analysis tools (for example, implemented as a ProM
plugins). The set of built-in functions that are used at
interpreting rules can be extended. As experiments,
the methods described in the related works
(Pegoraro, 2019; Kampik, 2022; Mitsyuk, 2017;
Marin-Castro, 2021; etc.) can be implemented with
created software.
REFERENCES
Abrosimova, P., Shalyaeva, I., Lyadova, L. (2018). The
Ontology-Based Event Mining Tools for Monitoring
Global Processes. In: Proceedings of the IEEE 12th
International Conference on Application of Information
and Communication Technologies (AICT 2018).
Almaty, 2018. Pp. 108-113.
Bano, D., Weske, M. (2020). Discovering Data Models
from Event Logs. In: Proc. International Conference on
Conceptual Modeling. Vienna, Austria, 2020. DOI:
10.1007/978-3-030-62522-1_5.
Bano, D., Zerbato, F., Weber, B. Weske, M. (2021).
Enhancing Discovered Process Models with Data
An Approach to Developing Ontology-Based Tools for Event Series Analysis
329

Object Lifecycles. In: Proc. IEEE 25th International
Enterprise Distributed Object Computing Conference
(EDOC). Gold Coast, Australia, 2021, pp. 124–133.
DOI: 10.1109/EDOC52215.2021.00023.
Calvanese, D., Montali, M., Syamsiyah, A.,
van der Aalst W.M.P. (2016). Ontology-Driven
Extraction of Event Logs from Relational Databases.
In: Business Process Management Workshops. BPM
2016. Lecture Notes in Business Information
Processing, vol 256. Springer, Cham.
Carrasquel, J.C., Chuburov, S.A., Lomazova, I.A. (2021).
Preprocessing Network Messages of Trading Systems
into Event Logs for Process Mining. In: Tools and
Methods of Program Analysis. TMPA 2019.
Communications in Computer and Information
Science, vol 1288. Springer, Cham. Pp. 88-100.
Cremerius, J., Weske, M. (2021). Data-Enhanced Process
Models in Process Mining. Preprint.
DOI: 10.48550/arXiv.2107.00565.
Cremerius, J., Weske, M. (2022). Change Detection in
Dynamic Event Attributes. In: Business Process
Management Forum. BPM 2022. Lecture Notes in
Business Information Processing. Vol. 458. Springer,
Cham. Pp. 157–172.
DOI: 10.1007/978-3-031-16171-1_10.
Cremerius, J., Weske, M. (2023). Context-Aware Change
Pattern Detection in Event Attributes of Recurring
Activities. In: Intelligent Information Systems. CAiSE
2023. Lecture Notes in Business Information
Processing. Vol. 477. Springer, Cham.
DOI: 10.1007/978-3-031-34674-3_1.
Kampik, T., Weske, M. (2022). Event Log Generation: An
Industry Perspective. In: Enterprise, Business-Process
and Information Systems Modeling. BPMDS EMMSAD
2022. Lecture Notes in Business Information
Processing. Vol. 450. Springer, Cham, pp. 123–136.
DOI: 10.1007/978-3-031-07475-2_9.
Kulagin, G., Ermakov, I., Lyadova, L. (2022). Ontology-
Based Development of Domain-Specific Languages via
Customizing Base Language. In: Proceedings of the
IEEE 16th International Conference on Application of
Information and Communication Technologies
(AICT2022). Washington DC, DC, USA: IEEE, 2022.
6 pp. DOI: 10.1109/AICT55583.2022.10013619.
Lanin, V., Lyadova, L., Zamyatina, E., Vostroknutov, N.
(2021). An Ontology-Based Approach to Social
Networks Mining. In: Proc. 13th International Joint
Conference on Knowledge Discovery, Knowledge
Engineering and Knowledge Management, 2021.
Vol. 2: KEOD. Lisbon : SciTePress, 2021.
Pp. 234-239. DOI: 10.5220/0010716600003064.
Lyadova, L., Suvorov, N., Zayakin, V., Zamyatina, E.
(2022). An Ontological Approach to the Development
of Analytical Platform Language Toolkits.
In: Proceedings of the IEEE 16th International
Conference on Application of Information and
Communication Technologies (AICT2022).
Washington DC, DC, USA: IEEE, 2022. 6 pp.
DOI: 10.1109/AICT55583.2022.10013576.
Marin-Castro, H.M., Tello-Leal, E. (2021). Event Log
Preprocessing for Process Mining: A Review. In:
Applied Sciences. 2021, 11(22), 10556. DOI:
10.3390/app112210556.
Mitsyuk, A.A., Shugurov, I.S., Kalenkova, A.A.,
van der Aalst, W. M.P. (2017). Generating event logs
for high-level process models. In: Simulation Modelling
Practice and Theory. Vol. 74, 2017, pp. 1–16.
DOI: 10.1016/j.simpat.2017.01.003.
Peña-Araya, V. (2015). Galean: Visualization of
Geolocated News Events from Social Media. In:
Proceedings of the 38th International ACM SIGIR
Conference on Research and Development in
Information Retrieval (SIGIR '15). ACM New York.
2015. Pp. 1041-1042.
Pegoraro, M., van der Aalst, W.M.P. (2019). Mining
Uncertain Event Data in Process Mining. In: Proc.
International Conference on Process Mining (ICPM).
Aachen, Germany, 2019, pp. 89–96. DOI:
10.1109/ICPM.2019.00023.
Ritter A. (2012). Open domain event extraction from
twitter. In: Proceedings of the 18th ACM SIGKDD
international conference on Knowledge discovery and
data mining. 2012. Pp. 1104-1112.
Shalyaeva, I., Lyadova, L., Lanin, V. (2016). Events
Analysis Based on Internet Information Retrieval and
Process Mining Tools. In: Proceedings of 10th
International Conference on Application of Information
and Communication Technologies (AICT2016). Baku,
2016, pp. 168-172.
Shalyaeva, I., Lyadova, L., Lanin, V. (2017). Ontology-
Driven System for Monitoring Global Processes on
Basis of Internet News. In: Proceedings of IEEE 11th
International Conference on Application of Information
and Communication Technologies (AICT2017).
Moscow, 2017, pp. 385-389.
Zayakin, V., Lyadova, L., Sminov, M., Lanin, V.,
Matta, N., Zamyatina, E. (2022). Event Series
Generation and Analysis Based on Multifaceted
Ontology. In: Proceedings of the IEEE 16th
International Conference on Application of Information
and Communication Technologies (AICT2022).
Washington DC, DC, USA: IEEE, 2022. 6 pp.
DOI: 10.1109/AICT55583.2022.10013573.
Zayakin, V.S., Lyadova, L.N., Lanin, V.V.,
Zamyatina, E.B., Rabchevskiy, E.A. (2023).
An Ontology-Driven Approach to the Analytical
Platform Development for Data-Intensive Domains. In:
Knowledge Discovery, Knowledge Engineering and
Knowledge Management. IC3K 2021. Communications
in Computer and Information Science. Vol. 1718, ch. 8,
pp. 129-149. Springer, Cham.
DOI: 10.1007/978-3-031-35924-8_8
KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development
330