REVIEWING PROCESS MINING APPLICATIONS AND TECHNIQUES IN EDUCATION

International Journal of Artificial Intelligence and Applications (IJAIA), Vol.13, No.1, January 2022
DOI: 10.5121/ijaia.2022.13106 83
REVIEWING PROCESS MINING APPLICATIONS
AND TECHNIQUES IN EDUCATION
Athanasios Sypsas and Dimitris Kalles
School of Science and Technology, Hellenic Open University, Patras, Greece
ABSTRACT
Process Mining (PM) emerged from business process management but has recently been applied to
educational data and has been found to facilitate the understanding of the educational process.
Educational Process Mining (EPM) bridges the gap between process analysis and data analysis, based on
the techniques of model discovery, conformance checking and extension of existing process models. We
present a systematic review of the recent and current status of research in the EPM domain, focusing on
application domains, techniques, tools and models, to highlight the use of EPM in comprehending and
improving educational processes.
KEYWORDS
Process Mining, Educational Process Mining, educational applications, process model
1. INTRODUCTION
The widespread use of information and communication systems in educational institutions and
training centers supports nearly all educational processes and has led to a vast amount of data
collection. Through extensive logging, these systems capture, apart from data, information
concerning the supported processes [1]. Data provide insightful information about learners’
assessment, learning resource usage, interaction logs, the processes they went through and their
involvement in the educational process in general [2], [3]. Additionally, the application of e-
learning and blended learning methods, provide helpful information about educational activities
at different levels of granularity [4]. Such collected data can be used by educators, learners,
institutions, or other educational stakeholders in order to analyze, evaluate and reveal different
aspects of the educational process.
Data mining is mainly used to extract useful information from gathered data in order to transform
it to a valuable pattern or structure for additional use [5]. Thus, the interest in using data mining
techniques in education is constantly increasing aiming to enhance the efficiency and quality of
teaching and learning. However, collected educational data are eventually required to be
understood and then transformed to actionable information, to achieve impact of some
significance [6]. The application of data mining techniques in educational data comes under the
term educational data mining (EDM). The main goal of this research area is to render most, if not
all, of such gathered data useful and valuable to the educational community, with a longer term
goal of offering a more personalized, adaptive and interactive educational environment for
learners [7].
A related area of research using mainly data mining methods is Learning Analytics (LA) [8]. LΑ
has the potential to detect and clarify characteristics of learning strategies by analyzing captured
data [9]. LA combines methods and techniques from other fields such as EDM, process mining

84
and technology enhanced learning. The generalized use of EDM and LA has led educational
institutions to use learning/ content management systems (LMS/ CMS) in order to collect data
related to improve the educational process to learners and learning processes [10]. Data mining
techniques, such as association analysis and classification have been successfully used with
various types of educational data i.e. LMS logs, assessment data, etc. [11]. Although data mining
can be used for getting a better understanding of educational processes and providing alternative
solutions to the interaction issues during educational process, most of the traditional data mining
techniques do not focus on the process as a whole, but on the result [4], [12]. When the focus of
the analysis is on the educational process, conventional EDM techniques do not normally lead to
meaningful results.
On a related front, process mining (PM) aims to discover, monitor, and improve processes by
extracting knowledge from event logs produced by information systems [13], [14]. Thus, PM
contributes to obtaining a meaningful process model from event logs helping identify possible
problematic issues concerning the investigated process [13]. Even though process mining has
emerged from the business community [15], it can also be applied on the EDM context [12]. So,
a new research field has emerged, with a focus on discovering and analyzing educational
processes, and it has been called Educational Process Mining (EPM) [5]. EPM focuses on the
events produced when a process is executed and which are related to completed activities. When
PM is applied in an EDM context, it contributes to knowledge and valuable information
extraction from educational event log files. Consequently, the research interest on EPM is
continuously increasing [2].
In general, a process in an educational context describes the necessary steps which will lead to a
specific learning outcome. For example, in virtual reality learning environments process models
can be used to abstractly describe the equipment of a physical environment and the experimental
processes carried out therein via simulation [16]. Therefore, EPM techniques, when applied on
these processes, facilitate the comprehension of experiments and learning flows, since the logged
learners’ actions are compared with the standardized procedures.
Even though PM is a relatively new technology, the potential of its application in the educational
domain is evident in recent research [2]. EPM, apart from data processing and analysis, facilitates
decision support and explores possible problematic situations during the educational process.
These situations concern, among others, time or performance bottlenecks and non-efficient
decision rules that affect the process model. The model extension as a EPM technique based on
decision mining, performance analysis and user profiling contributes to detecting and effectively
dealing with such situations [3]. Τhe three main types of process mining when applied an
educational context are [4], [17]:
1. Process model discovery. The process model is constructed from event logs.
2. Conformance checking. The model is compared to the event log and deviations are detected.
3. Process model extension. The process model is extended or improved using information from
the event logs.
All three aforementioned approaches cannot be implemented using traditional EDM techniques.
So, new tools and methods were developed in order to extract process-related knowledge from
the produced event logs. The application of process mining in the educational field [11] is
illustrated in Figure 1 and we briefly review its life cycle below.
The educational information system (i.e., LMS) produces and stores event logs (traces from
executed processes). Such events could be activity logs in virtual educational environments
(virtual laboratories), learners’ registration procedures, access to educational material logs,

85
Massive Open Online Courses (MOCs), Intelligent Tutoring Systems (ITSs), etc. Using process
mining techniques event logs are associated with the process models leading to the extraction of
the appropriate information and knowledge concerning process checking, possible execution
problems discovering and errors prediction.
Figure 1. Process mining application in the educational field [23]
However, successful knowledge extraction depends on the quality of the stored event logs, since
the huge volume and the heterogeneity of possible traces poses challenges for process mining
techniques and algorithms [3], [18], [19]. Using EPM the underlying educational processes can
be better understood leading to recommendations and advice to learners, feedback generation for
learners, educators and researchers, detecting specific learning disabilities, and improving
management of learning objects [20]. The approaches in the present review concern either some
or all process mining approaches (Process model discovery, Conformance checking, Process
model extension), depending on the application focus and scope.
In this survey, we focus mainly on exploring the various applications of process mining
techniques in the education field and, especially, in virtual environments. We aim to provide a
comprehensive overview distilling the state-of-the-art from scientific journals and conference
proceedings so as to assist researchers assessing recent progress in educational process mining.
We have, thus, structured the rest of this paper as follows: In Section 2 the previous work in EPM
is summarized, while in Section 3 an overview of research methodology along with research
questions is presented. Section 4 outlines the findings with respect to the research questions
defined in Section 3 and the final Section 5 provides the conclusion.
2. PREVIOUS WORKS ON EPM
Research on educational process mining is not a new topic. In fact, several reviews have been
published concerning process mining in general [13], [17], [21]. However, these reviews do not
focus on EPM. Only two other reviews were found that discuss EPM more in depth [2], [22].
Bogarín et al. [2] published a survey on the subject introducing the EPM and elaborating on some
of the potential of the specific research area. They presented the components of an EPM
framework, the data, tools, techniques and models used in EPM. The papers included were
mainly categorized by educational application domains. The major drawback was that it was not

86
a systematic literature review and the included papers are published by year 2017. Ghazal et al.
[22] conducted a literature review collecting 37 papers, containing case studies where EPM was
applied. The authors searched for tools and techniques used in EPM and analyzed the selected
case studies and their results. Furthermore, they have discovered and highlighted the potential
benefits of EPM. However, results concern case studies by the year 2016.
Our work is distinct from these literature reviews, since the results are up-to-date and includes
application domains that were not covered before. From our research was concluded that papers
published after 2017 are almost the double compared to the papers before 2017. Finally, this
study considers research on all aspects of EPM application, including not only educational case
studies, but also administration and curriculum issues.
3. RESEARCH APPROACH
This section covers the research approach used in order to obtain as much relevant information as
possible concerning the EPM applications. This study employed literature review to extract the
various applications in the research field of EPM.
3.1. Research Questions
The key part of a systematic review is the definition of research questions [23]. In the present
review paper, the defined research questions contribute in building a knowledge base about the
current state of concepts.
RQ1: Where is Educational Process Mining applied?
RQ2: What are the techniques and/or algorithms used in Educational Process Mining research?
RQ3: How is EPM applied on Virtual educational environments?
To answer the aforementioned research questions a search was conducted.
3.2. Conduct Search
In order to execute a literature review, a broad search plan is constructed, so as to ensure the
completeness of the search of resources under consideration. A precise search query is needed for
scientific databases search, taking into account the related terms extracted from research
questions, as can be seen in Table 1, below.
Table 1. Search terms
Key term Broader terms Priority
Process mining Educational process mining, data mining 1
Education - 2
Techniques Algorithms 3
Domain Virtual environment, virtual laboratory,
case study
4
Based on the search terms of Table 1, various search strings were tested. After numerous
iterations the search string that resulted in the highest number of papers is the following:
(“Process mining” AND (“education” OR “virtual laboratory” OR “virtual environment”) AND
(“case study” OR “domain” OR “techniques”)). The main source of information and data was the
peer-reviewed literature. Searches for peer-reviewed journal articles and dissertations were
conducted using the following databases: ACM Digital Library, SCOPUS, Spinger Link and

87
ScienceDirect. Moreover, sources outside those databases were searched, particularly using the
references of relevant articles. Google Scholar was used, but the overlapped articles with other
data sources were excluded. The search was performed in June 2021. All the web searches were
executed without cookies or login to guarantee any impact avoidance due to previous searches.
After the initial search, filtering options offered by digital databases were used in order to apply
inclusion (IC) and exclusion (EC) criteria. The inclusion criteria were:
• Peer-reviewed articles published in journals, International Conferences/Workshop
Proceedings;
• Full results presentation;
• Frequency of citations;
• Thorough methodology presentation;
• English language researches.
Subsequently, the exclusion criteria were applied on the results. These criteria were:
• Papers not written in English;
• Duplicate papers;
• Editorials or interviews;
• Very short papers, one or two pages long;
• Missing authors papers.
After applying IC and EC 832 articles were remained for review. Publications that obviously
were out of the interests of the current study were removed. Then the evaluation focused on
abstracts: they were manually filtered by two researchers as relevant to the research criteria and
worthy of further study. For the 112 publications that were relevant to our literature search, the
full articles were accessed. In the case of any uncertainty concerning a study’s relevance based on
its title or abstract, we chose to include the study in our full-text evaluation. At the final stage, the
full texts of these 112 articles were examined for final worthiness. The resulting literature
revealed 95 peer reviewed journal articles and documents. Each document was listed to
categorize the information for further analysis, documents were grouped, summarized and
critically analyzed.
4. FINDINGS
This section presents findings with respect to the RQs defined in section 3.1.
4.1. Educational Process Mining Application Domains
A wide range of EPM applications in various educational domains is observed. In this subsection
the main body of the reviewed literature is presented, giving special significance to the EPM
applications and their results.
4.1.1. LMS, LAD and Online courses
Learning Management System (LMS) and Learning Analytics Dashboard (LAD) are online
learning environments that offer various learning opportunities and data concerning learning
activities. There is a constantly increasing research about the application of EPM in this domain,
since these learning environments are used in many educational settings.

88
In [24] PM techniques were applied to analyse LMS logs from Moodle in order to examine the
alignment of teacher's plan and students' use of LMS resources. The results revealed that
students’ behavior did not fit the plan teachers had in mind. In a recent study [25] institutional
Moodle was used as input for PM application and an ad-hoc tool was developed. LMS was also
used in [26] to detect effective learning strategies. Using PM techniques researchers concluded
that their method could activate learners’ awareness about effective learning strategies and
motivate them to use them productively. Combining clustering and sequential pattern mining,
Martínez-Carrascal et al. in [27], applied generalized sequential pattern (GSP) algorithm on the
log data. This technique aimed to detect differences between two groups of students (passing and
failing). In an earlier research [28] Moodle was used in order to assess the LMS usage behavior.
Authors concluded that process model can be used to tailor the Moodle LMS to the specific needs
of the individual courses. The same technique was used in [29]. In [30] dashboards of 35 school
teachers were retrieved. PM techniques were used and outcomes showed that teachers paid
attention on the group alerts indicating support need. Etinger [31] applied PM techniques in order
to discover LMS Course Usage Patterns. Using a new PM algorithm, Inductive Miner (IM),
Bogarín et al. [32] tried to discover learning processes under LMS Moodle. As a result, learning
model discovery with IM could be a resource for preventing learning failure in LMSs. In a
similar environment Ahmad Uzir et al. [33] proposed the use of LMS records to discover time
management strategies under self-regulated learning (SRL) approach. Real et al. [34] used PM to
verify student learning paths in an Introductory Programming Course. The event logs were
collected in the Moodle and the different learning paths were revealed. Along similar lines,
Romero et al. [5] proposed the adoption of clustering to improve EPM. They used Moodle log
files to optimize both the fitness and comprehensibility of the model providing useful information
to educators about student learning and the failing course risk. Bogarín et al. [35] based on data
from LMS logs managed to obtain more specific and accurate models of students' bahaviour. The
online modeling platform BJUTModeling was used in studies [36] and [37] as a tool for student
learning evaluation effect. Based on PM, authors achieved student modeling patterns mining and
their behaviors interpretation. In an early study [38], PM contributed in exploring regulatory
processes under Moodle. Similarly, in [39] Moodle log files were used and PM was proved to be
helpful for tutors to effectively cope with the study process of their students. In a study [40] with
1.500 participants in distance learning domain, event logs generated by LMS served as input for
PM analysis. Results suggested that applying either classical process mining or modern data
mining techniques could offer valuable benefits for this domain.
Cerezo et al. [12] explored students’ SRL processes, during an online course by using PM
Techniques. They managed to discover models with good fitness values, meaning that students’
interactions on the Moodle platform were correctly reproduced. Based on Moodle log files
Pechenizkiy et al. [11] applied process mining techniques to online assessment data, illustrated
some of the potential PM techniques offer. Likewise, in [41] PM was used to understand
students’ approaches to learning programming. They used a combination of methods (data from
more than one source, process and sequence mining) to gain an in-depth understanding of the
programming learning. Shabaninejad, et al. [42] used LAD as source of rich and complex data
about students and their learning activities. The learning processes of students were visualized
and compared giving insights to educators that could be used to adjust educational process. In a
recent research [43] a method was proposed to discover student’s activities, navigation paths, and
behavior in LMS Moodle. PM application offered the ability to distinguish different variants
(patterns) of student activity, helping educators to detect problematic issues during learning
process. The same LMS was selected for another study [44], in order to explore student behavior.
Specifically, students’ interactions while engaging in specific quiz-based learning activities were
analyzed contributing in research about students’ quiz-taking strategies in LMS. In an online
course on system dynamics Reimann et al. [45] used PM to discover models from interaction data
in LMS.

89
4.1.2. MOOCs
Massive Open Online Courses offer free, at most cases, learning opportunities to the online
community. Due to the big number of participants, the produced log files serve as input for PM
techniques leading to valuable research on the area.
The research [46] involved log data from 16.224 learners, participated in a MOOC. Applying PM
and clustering techniques researchers examined the relation between learning behavior and
learning progress, offering guidance for educators. In another study [9] PM was used on the trace
data originating from learners’ interactions with the learning activities in a MOOC to explore the
association of learning strategies and the learners’ personality traits. Bernal et al. [47] in a MOOC
under Coursera platform used PM on activity data from the MOOC log files to extend the
knowledge about students’ behavior in online environments. In order to study dropout prediction
in MOOCs process and sequence mining techniques were applied by Deeva et al. [48] on event
data gathered from the edX platform. EdX platform was also used to study educational process
intelligence (EPI) [49] providing a contextual-based and informed learning process analysis and
decision making through the PM approach. A FutureLearn MOOC was utilized as data resource,
in order to investigate variation in learning processes [50]. PM discovery method showed that
insights from learners’ behavior can inform MOOC designers about adjusting learning activities
to different groups of learners based on their goals. Three MOOCs offered resources for
exploration, through PM application, of the relationship between theoretical SRL strategies and
behavioural patterns on MOOC platforms [51]. Finally, Rizvi et al. [52] worked with logged data
from an Open University MOOC offered via FutureLearn. The study compared behaviors
concerning participation of both groups with expected learning behavior for all types of learning
activities.
4.1.3. E-learning Environments, Micro-Learning Environments, Intelligent Learning
Environments and Intelligent Tutoring Environments
Various e-leaning tools have been designed and developed for learning activities. Beheshitha, et
al. [53] proposed the application of nStudy, a learning tool offered as a Firefox add-on, in order
to capture log data of learners’ activities within the environment. PM was used to examine the
relationship between students' self-reported aptitudes and how they self-regulate their learning in
action. A computer-based context, nBrowser, was applied in a recent study [54] to examine how
teachers enact SRL as they move from the beginning to later stages, by applying PM on logged
data. Okoye et al. [55] applied PM to discover sets of recurrent behaviors that can be found
within a learning process. Using activity logs within the learning process, the authors determined
suitable learning patterns by means of semantic reasoning which can then be used to address the
problem of adapting learning to the captured user profiles.
Students’ work in Open-ended Learning Environments (OELEs) derives a holistic view of
students’ learning behaviors. In [56] PM was applied on logged actions from Betty’s Brain
learning environment and discovered that high performing students not only execute more
actions, but their actions are also supported and effective. Intelligent Learning Environments
(ILEs) offer useful data for learning behaviors. So, Jeong et al., [57] in a rather early study, took
advantage of them to analyze, via PM, the differences in transitionary state behavior between the
high and the low- performing learners. In a hypermedia learning environment, serving as a
computer-based learning environment (CBLE), process model discovery and conformance
checking were used to investigate both the effect of metacognitive prompts on metacognitive and
cognitive learning processes and the SRL process itself [58]. An e-Learning environment (DME)
for software operation data offered the log framework XES as input for PM technique of
conformance checking [59]. Authors reported that defining and obtaining the right software

90
operation data is a process and PM contributed in process analysis. Deeva et al. [48] analyzed
behavioral data from the JMermaid modeling environment to explore potential behavioral
patterns and indicative features correlated with better learning outcomes. The same modeling
environment was selected for another study [60]. The interactions with the modeling tool have
been logged and PM was used to examine how educators endorse SRL as they move from the
beginning to later stages of designing lessons. An intelligent tutoring system offered input data
for PM application in a research [61], scoping to characterize learner behaviors at different
phases of problem solving. Deeva and Weerdt [62] used a Smart Learning Environment. Various
PM techniques were applied leading to student interactions capture in order to improve learning
experiences. Poncin et al. [63] made use of PM techniques on Capstone projects educational
approach to obtain insights in the development process followed. In a study [64] a cloud-based
Educational Software Tool was developed to offer resources for PM analysis aiming at
understanding and validating the behavior of team projects in capstone courses. Micro-Learning
is a new leaning approach based on e-mail or SMS, chat rooms, microblogging in which a
primary learning resource is split in a series of micro learning units that are distributed via the
Internet. Chen et al. [65] proposed the application of PM to support personalized micro-learning
by spotting and extracting the suitable micro- learning units and then organize them as a suitable
learning process matching the learners’ needs.
A hypermedia learning environment worked as the data collector for a study [66]. Model
discovery and conformance checking were combined to analyze the temporal order of
spontaneous individual regulation activities and how findings can be applied to identify process
patterns in SRL events. Hypermedia learning sessions were used in [67], applying PM to detect
long-term effects of metacognitive prompting on SRL in a follow-up learning task without
instructional support. Combining the Interactive Digital Narratives (IDN) approach under
Prolific, a scientific crowdsourcing platform and PM, Estupiñán and Szilas [68] examined the
understanding of user engagement, particularly, spotting when and what happened when user
engagement dropped.
4.1.4. Flipped Classroom approaches and blended learning approaches
The combination of e-learning and face-to-face educational approaches offers a wide range of
applications resulting in learner- friendly environments. These approaches are utilized to extract
valuable information when PM techniques are applied.
In a recent study [69] researchers proposed PM usage in event logs derived from project-based,
collaborative undergraduate engineering course to illuminate how regulation phases can facilitate
subsequent regulation. In [70] a three-year study was conducted, applying PM techniques on the
data collected from class preparation activities of a flipped classroom targeting to reveal useful
information from trace data and the way they are associated with academic performance and
feedback. Similarly, in another study [71] concerning a course that followed a flipped classroom
pedagogy to study SRL, PM was used for a richer comprehension into SRL behaviors. The
application of PM at a flipped classroom approach aiming at providing real-time or after-class
indicators to students and educators was studied in [72]. At a similar flipped classroom approach
Saint, Gašević et al. [73] discussed the PM use for learning strategies detection, concluding that
more strategically minded activity, embodying learner SRL, generally proves to be more
successful than less disciplined reactive behaviors. Likewise, in [74] the PM application to
support SRL under flipped classroom environment was explored. Bolt et al [75] combined the
complementary approaches of process cubes and analytic workflows in the field of PM to analyze
large, heterogeneous event data of multidimensional nature. Logged data offered resources for an
older study [76], where PM application implemented the comparison of the learning processes
inside and between courses contributing in better understanding of the study behavior of students.

91
Heirweg et al. [77] contrary to the very limited existing research in this regard, applied PM on
online data collection through thinking aloud method exploring the sequence in which SRL
activities are conducted during learning. Under a blended-learning approach PM was used to
identify process patterns as part of the tailored feedback on medical procedural training [78].
Dahari et al. [79]extended a PM tool (ProM) framework for educational purposes in order to
achieve a rich development environment for executing, creating, and extending mining
techniques.
4.1.5. Online Collaborative Environments
Online collaborative environments such as wikis and social networks pose a challenge for PM
application, since the mined data and logged events can be easily capture for further analysis.
Virtual collaborative system KP-Lab System is used in an early study [80] where a standalone
tool served for PM application, via critical patterns recognition, on the data logged by the system
aiming to study learners’ activities. Ten years later Chen and Wang [81] studied the use of a
Collaborative Web-based Inquiry Science Environment (CWISE) as data collector for PM
application in order to study effective learning behaviors. In the same way [82] the PM applied
on event logs collected and captured during a distance learning activity in a computer-based
collaborative learning environment.
Social networks via group formation offer an online collaborative space for educational
processes. Thus, Peeters et al. [83] exploited Facebook as an online collaborative environment for
foreign language learners. Through PM the researchers analyzed a range of SRL tactics and task
phases followed by learners to manage their academic writing process. Another social network
(WhatsApp) served as the collaboration platform in [84] where PM techniques included data pre-
processing, discovering process, and conformance checking provided perceptions that can be
used to improve student collaboration.
Although in recent years wikis are not used frequently, they can serve as a repository for learning
activities. Hence, in an early study [85] PM was applied on activity logs generated by process
aware information systems such as Wiki based document management system, to help the
educator in giving directed and specific feedback. Caballero-Hernández et al. [86] suggested the
use of PM techniques on wiki log data to analyse the behavior of team members in sequential
processes like collaborative projects.
4.1.6. Online collaborative environments
Game Based Learning (GBL) is the educational approach where serious games are used for
educational purposes. The main purpose of entertainment games is the fun of the player while
serious games have educational purposes. Hernández [87] used the PM technique of event
sequence analysis to study skill assessment in learning experiences based on serious games. PM
application on events in MySQL database of a game-based assessment tool, Shadowspect,
revealed how students interacted with the game and spotted the problems that the students were
having while playing [27]. Authors in [88] used prototype metrics to process and analyze
students’ self-reflection in the context of Radix, a multi-player inquiry-based learning game on
STEM topics scoping to expand the game itself into a more complete game-based learning
system.
Finally, a research [89] explored the possibilities of application of PM techniques in the analysis
of interactions among players, and the involved decision taking processes, in non-digital games.

92
4.1.7. Curriculum and carrier analysis
One of the important steps to ensure that a syllabus aligns with the educational setting’s mission
as well as the education standards from the government is curriculum assessment. When
curriculum model is taken under consideration, students have a guideline to consider the learning
process during their study period [90]. Having this in mind, the abovementioned researchers
aimed at investigating the actual curriculum model by constructing the model from the students’
academic results using PM techniques. Based on the curriculum analysis when sequence
matching analysis via conformance checking is performed, another study [91] evaluated the a-
prori curriculum design with actual curriculum learning sequence. In one of the earliest studies
[92] a new domain-driven EPM framework was proposed to facilitate more focused search for
local patterns based on logged data from academic curriculum.
As far as carrier analysis is concerned, Cameranesi et al. [93] utilized the set of traces
representing students’ careers from ESSE3, a software suite commonly adopted by Italian
Universities to manage students’ careers and performed PM application that led to a preliminary
analysis of students' careers.
4.1.8. Students’ enrolment and performance assessment
A long-term research which aimed at studying student’s enrolment used data was based on
anonymized enrolment and course-grade records for 6 cohorts of students between 2004 and
2009 [94]. PM discovery algorithms, combined with statistical techniques, provided visual
representations that allow an easy-to-understand analysis, helping to comprehend the process that
led undergraduate students to drop out. Salazar-Fernandez et al. [95] made use of the same
educational trajectories of engineering students in each one of the high-failure rate courses as the
previous study. PM analysis of trajectories allowed researchers to confirm the importance of prior
time and resources investment regarding the students’ decision to continue their studies. A data
set with relevant information about the performance of 612 students was the source for PM
application in [96]. The analysis of students’ performance revealed courses which possibly
represent bottlenecks in the process of completing a computer science degree.
Online Assessment Test is an evaluation form of a student learning process. Aisa et al. [97]
evaluated the online assessment test via PM application so as to extract the frequency of each
task ID revealing the strengths and weaknesses of each student. In a similar study, workplace
based assessment (WBA) was suggested as a tool to collect feedback on students’ practical skills,
their performance, and goals setting [98].
4.1.9. Professional training
Many organizations and companies invest capitals on their employees training targeting at their
professional development. From our research five studies exploited the PM application for
professional training. Ariouat et al. [20] suggested a two-step clustering PM technique on real-
world professional training databases from a worldwide consulting company. Their approach was
adequate to partition efficiently training event logs in order to achieve a better understanding of
the underlying educational processes. In a 6.000 employees study [99], semantic annotations to
educational event logs were associated. Thus, more accurate and compact educational processes
were extracted and analyzed at different levels of abstraction providing useful information for
employees’ professional training. Cairns et al. [100] used a process log, issued from the Altran
training institute activities. Authors used sequence clustering and then PM techniques which
contributed in the development of a distributed platform dedicated to educational process model
discovery and conformance checking.

93
In Medical Training domain, a recent research [78] studied the PM application on surgical
procedures. As a result, tailored feedback of surgical procedures using a process perspective was
mined. In the same domain the basic concepts to design medical training applications using PM
was introduced [101]. The proposed methodology Process-Oriented Medical Education (POME)
was applied on a running case of a surgical procedure, which has been used to develop
applications in medical training using PM.
4.1.10. Educational Virtual Environments
Educational Virtual Environments are offering an educational approach, where learners interact
with educators or with other learners in a virtual environment. Additionally, in some cases, they
implement a virtual world where students are conducting experiments which could not be
executed in a real life setting i.e., not accessible chemistry or biology laboratories. In such
environments, PM has been used to analyze the learning processes taking place there. Sypsas and
Kalles [102] proposed the application of conformance checking PM technique on logged events
from a virtual biology laboratory (Onlabs). The authors suggested that the research outcomes
assist educators to decide whether the specific virtual lab can be used as an educational tool in
various settings i.e., universities, secondary education schools, etc. A 3D biotechnology virtual
lab combined with an Intelligent Tutoring System (ITS) is used in another study [103]. PM on
student logs were used to implement an automatic tutoring adapted to each type of student by
applying methods of extraction and analysis of data. Two other studies presented the PM usage
on data taken out from logged events inside virtual learning environments [104], [105]. Units of
learning (UoLs) from different domains have been used to extract the learning flow structure and
to obtain the underlying rules that control the adaptive learning of students. Proposed approaches
claimed that facilitated the reuse of UoLs defined in virtual learning environments. Castillo [106]
suggested the PM use in the context of a virtual laboratory for a multiagent programming course
to detect successful and failed behaviors of students. Another simulation-based environment
TELEOS (Technology Enhanced Learning Environment for Orthopedic Surgery) served as a data
source in [107]. The extraction of sequential rules common to several sequences was used to
mine the distinctive phases of simulated operations of vertebroplasty.
Finally, Fernández-Gallego et al. [108] used an Opensim-based virtual world platform, known as
OPENET4VE, to develop applications that automatically extract the learning flow undertaken by
students in virtual worlds.
4.1.11. Real World Processes
In [109] a novel PM technique for linguistic summarization of event logs was suggested, which
generated linguistic summaries that are concise enough to be used in a practical setting.
Execution log of a Dutch municipality process was utilized and results were proved to be
practically usable by a human interpreter.
In conclusion, in Figure 2 the number of each application is presented. The most active domains
are LMS, LAD and Online courses, E-learning Environments, Micro-Learning Environments,
Intelligent Learning Environments and Intelligent Tutoring Environments, MOOCs and Flipped
Classroom and blended learning approaches. Some studies (4) can be classified in two domains
as they use approaches from more than one.

94
Figure 2. EPM studies per domain
4.2. Techniques and/or Algorithms used in Educational Process Mining
Applications
Most studies in the present paper used PM techniques and algorithms that have already been
implemented in PM tools such as ProM and Disco. However, some researchers combined
techniques and algorithms from Data Mining and Learning Analytics domains with PM in order
to achieve better results. Additionally, they developed plug-ins in existing tools or new tools
which implement these approaches. In Figure 3 below, the PM tools are presented, showing that
the open-source tool ProM and the Disco (Fluxicon) tool dominate in EPM applications.
Figure 3. PM tools used in EPM studies
The model discovery and conformance checking techniques are used almost in all studies. In a
rather big number of studies more than one algorithm and techniques are used targeting at finding
the optimal solution. Figure 4 depicts the algorithms and techniques that are used in the studies
included in our research. The most commonly used algorithms are Fuzzy miner (23.62%),

95
Heuristic miner (18.89%) and Inductive miner (11.02%), whistle Conformance checking is
adopted by 13 researches.
Figure 4. PM algorithms and techniques used in EPM
Model discovery and conformance checking are using model representation. The most used
representations from the studies where process map (21,11%), Petri nets (17,78%), process model
(10%) and various kinds of Graphs (cyclic, directed, weighted, etc.) (10%).
4.3. EPM on Virtual Educational Environments
Virtual educational environments have been adopted as educational tools during the last decade,
since they offer an attractive and interactive learning environment. The most used applications on
the domain are the virtual laboratories and the virtual worlds. In these applications process
models are used to represent both the virtual learning environment i.e., laboratory equipment and
the learners’ steps, i.e., experimental steps. The seven studies using EPM in virtual educational
environments are presented in section 4.1.10. Four out of seven studies concern virtual
laboratories, two virtual laboratories for biology, biotechnology, one for medical training and one
for computer science. Two studies are based on Virtual Learning Environments and one on
virtual worlds. EPM techniques applied are process discovery and conformance checking based
on the log files produced by the virtual environments. Four papers used Petri nets for model
representation, while the other three Bayesian network, extended automata and process map. As
far as EPM tools are concerned the majority (57, 15%) of researches adopted prototype tools
even though there are free tools available (ProM).

96
5. CONCLUSIONS
Even though PM was firstly applied in more traditional business domain, the lack of a process-
centric approach concerning educational data has spearheaded work on exploring the
applicability of PM to the education domain, leading to the emergence of EPM [110]. The present
research review introduces EPM and provides an overview of its applications in various
educational domains offering researchers, educators and institutions a view of the different
approaches over the last decade. Additionally, tools, techniques and algorithms are presented to
contribute to a deeper comprehension of EPM and its applications. The analysis of the selected
studies exemplified the approaches and tools used by researchers, and highlighted the potential
benefits of EPM. Most applications concern online courses and e-learning environments, since
the used tools (LMS) offer logging capabilities for learners’ data and procedures. Furthermore,
many papers suggest the use of EPM combined with SRL leading to learners’ behavior study.
Thus, educators and institutions are able to detect possible problematic situations during
educational process, mainly in distance education courses, where SRL is a prerequisite.
EPM, as an evolving research area, faces many challenges for the future. A first one is the
improvement of model discovery, since the model is used to abstractly describe all possible
learners’ actions and processes taking place during the educational procedure. Thus, improved
model discovery may capture as many processes as possible for further EPM analysis. Secondly,
semantics could be used to provide a more theoretical analysis of the processes taking place in
educational procedure. This may lead to the design of more detailed log generation application in
the existing information systems as semantics are associated with the logged events. This
approached was also mentioned in a study [3] where labels were suggested to be linked with
events. Thirdly, the results of EPM application should be formed in order to be understandable
from all the involved parts, educators, learners, institutions, stakeholders. This, could serve as an
initiative for the development of interactive environments with smart visualization techniques
contributing on a more effective and efficient use of EPM [3]. Lastly, EPM tools and EPM
datasets could be freely available so that researchers can use them to test their approaches
offering a boost on EPM research.
As far as virtual laboratories are concerned, the potential of EPM applications is not fully
explored. Irritative challenges among others are the implementation of applications that
efficiently log processes from as many experiments from various domains as possible and the
detection of learners’ actions inside the lab that need improvement compared with learning
outcomes that have been set. This could be achieved through the incorporation of a virtual lab
assistant or chatbot.
This review aims at providing the educational community and researchers with the up-to-date
EPM techniques and applications on educational data. EPM facilitates a better comprehension of
the underlying educational process from raw event data contributing in effective learning results.
However, special attention should be paid on AI applications, as EU guidelines
(https://ec.europa.eu/commission/presscorner/detail/en/ip_21_1682) identify AI systems as high-
risk when used in Educational or vocational training, as it may determine the access to education
and professional course of someone's life (e.g., scoring of exams).
REFERENCES
[1] A. A. Kalenkova and D. A. Kolesnikov, “Application of a Genetic Algorithm for Finding Edit
Distances between Process Models,” Autom. Control Comput. Sci., vol. 53, no. 7, pp. 617–627, 2019.
[2] A. Bogarín, R. Cerezo, and C. Romero, “A survey on educational process mining,” Wiley Interdiscip.
Rev. Data Min. Knowl. Discov., vol. 8, no. 1, p. e1230, 2018.

97
[3] A. H. Cairns, B. Gueni, M. Fhima, A. Cairns, S. David, and N. Khelifa, “Process Mining in the
Education Domain,” Int. J. Adv. Intell. Syst., vol. 08, no. 1 & 2, pp. 219–232, 2015.
[4] N. Trčka, M. Pechenizkiy, and W. van der Aalst, “Process mining from educational data,” in
Handbook of Educational Data Mining, Chapman &., C. Romero, S. Ventura, M. Pechenizkiy, and
R. Baker, Eds. CRC Press, 2010, pp. 123–142.
[5] C. Romero, R. Cerezo, A. Bogarín, and M. Sánchez-Santillán, “EDUCATIONAL PROCESS
MINING: A TUTORIAL AND CASE STUDY USING MOODLE DATA SETS,” in Data Mining
and Learning Analytics: Applications in Educational Research, S. ELATIA, D. IPPERCIEL, and O.
R. ZAÏANE, Eds. John Wiley & Sons, Inc., 2016, pp. 3–28.
[6] R. Tsoni, E. C. Stavropoulos, and V. S. Verykios, “Leveraging Learning Analytics with the Power of
Words,” Envisioning Rep. Empower. Univ., vol. 24, pp. 24–26, 2019.
[7] H. Aldowah, H. Al-Samarraie, and W. M. Fauzy, “Educational data mining and learning analytics for
21st century higher education: A review and synthesis,” Telemat. Informatics, vol. 37, no. January,
pp. 13–49, 2019.
[8] K. H. Kyritsi, V. Zorkadis, E. C. Stavropoulos, and V. S. Verykios, “The pursuit of patterns in
educational data mining as a threat to student privacy,” J. Interact. Media Educ., vol. 2019, no. 1, pp.
1–10, 2019.
[9] W. Matcha et al., “Analytics of learning strategies: The association with the personality traits,” ACM
Int. Conf. Proceeding Ser., pp. 151–160, 2020.
[10] M. M. Tair and A. M. El-Halees, “Mining Educational Data to Improve Students’ Performance: A
Case Study,” Int. J. Inf. Commun. Technol. Res., vol. 2, no. 2, pp. 140–146, 2012.
[11] M. Pechenizkiy, N. Trčka, E. Vasilyeva, W. Van Der Aalst, and P. De Bra, “Process mining online
assessment data,” in EDM’09 - Educational Data Mining 2009: 2nd International Conference on
Educational Data Mining, 2009, no. June 2014, pp. 279–288.
[12] R. Cerezo, A. Bogarín, M. Esteban, and C. Romero, “Process mining for self-regulated learning
assessment in e-learning,” J. Comput. High. Educ., vol. 32, no. 1, pp. 74–88, 2020.
[13] C. dos S. Garcia et al., “Process mining techniques and applications – A systematic mapping study,”
Expert Syst. Appl., vol. 133, pp. 260–295, 2019.
[14] W. Van Der Aalst, “Process mining,” Commun. ACM, vol. 55, no. 8, pp. 76–83, 2012.
[15] W. Van Der Aalst, T. Weijters, and L. Maruster, “Workflow mining: Discovering process models
from event logs,” IEEE Trans. Knowl. Data Eng., vol. 16, no. 9, pp. 1128–1142, 2004.
[16] J. A. Rossiter, “Low production cost virtual modelling and control laboratories for chemical
engineering students,” IFAC-PapersOnLine, vol. 49, no. 6, pp. 230–235, Jan. 2016.
[17] E. Rojas, J. Munoz-Gama, M. Sepúlveda, and D. Capurro, “Process mining in healthcare: A literature
review,” J. Biomed. Inform., vol. 61, pp. 224–236, 2016.
[18] R. P. J. C. Bose, R. S. Mans, W. M. P. Van Der Aalst, B. F. A. Hompes, H. M. W. Verbeek, and W.
M. P. Vanderaalst, “Wanna improve process mining results?,” Proc. 2013 IEEE Symp. Comput.
Intell. Data Mining, CIDM 2013 - 2013 IEEE Symp. Ser. Comput. Intell. SSCI 2013, vol. 237, pp. 32–
57, 2015.
[19] B. F. A. Hompes, H. M. W. Verbeek, and W. M. P. Van deraalst, “Finding suitable activity clusters
for decomposed process discovery,” in Lecture Notes in Business Information Processing, 2015, vol.
237, pp. 32–57.
[20] H. Ariouat, A. H. Cairns, K. Barkaoui, J. Akoka, and N. Khelifa, “A two-step clustering approach for
improving educational process model discovery,” Proc. - 25th IEEE Int. Conf. Enabling Technol.
Infrastruct. Collab. Enterp. WETICE 2016, no. June, pp. 38–43, 2016.
[21] D. Dakic, D. Stefanovic, T. Lolic, D. Narandzic, and N. Simeunovic, “Event Log Extraction for the
Purpose of Process Mining: A Systematic Literature Review,” in 15th International Symposium in
Management, 2020, no. October, pp. 299–312.
[22] M. A. Ghazal, O. Ibrahim, and M. A. Salama, “Educational process mining: A systematic literature
review,” in Proceedings - 2017 European Conference on Electrical Engineering and Computer
Science, EECS 2017, 2018, pp. 198–203.
[23] B. Kitchenham, O. Pearl Brereton, D. Budgen, M. Turner, J. Bailey, and S. Linkman, “Systematic
literature reviews in software engineering - A systematic literature review,” Inf. Softw. Technol., vol.
51, no. 1, pp. 7–15, 2009.
[24] P. Alvarez, J. Fabra, S. Hernandez, and J. Ezpeleta, “Alignment of teacher’s plan and students’ use of
LMS resources. Analysis of Moodle logs,” 2016 15th Int. Conf. Inf. Technol. Based High. Educ.
Training, ITHET 2016, 2016.

98
[25] A. Balderas and J. A. Caballero-Hernández, “Analysis of Learning Records to Detect Student
Cheating on Online Exams: Case Study during COVID-19 Pandemic,” in Eighth International
Conference on Technological Ecosystems for Enhancing Multiculturality (TEEM’20), 2020, pp. 752–
757.
[26] N. A. A. Uzir, D. Gaševic, J. Jovanovic, W. Matcha, L. A. Lim, and A. Fudge, “Analytics of time
management and learning strategies for effective online learning in blended environments,” ACM Int.
Conf. Proceeding Ser., pp. 392–401, 2020.
[27] M. J. Gomez, J. A. Ruipérez-Valiente, P. A. Martinez, and Y. J. Kim, “Exploring the Affordances of
Sequence Mining in Educational Games,” in TEEM’20: Eighth International Conference on
Technological Ecosystems for Enhancing Multiculturality, 2020, pp. 648–654.
[28] D. Etinger, T. Orehovački, and S. Babić, “Applying process mining techniques to learning
management systems for educational process model discovery and analysis,” in Advances in
Intelligent Systems and Computing, 2018, vol. 722, pp. 420–425.
[29] N. Kelly et al., “Combining event- and variable- centred approaches to institution- facing learning
analytics at the unit of study level,” Int. J. Inf. Learn. Technol. Artic. Inf., vol. 34, no. 1, pp. 63–78,
2017.
[30] A. Van Leeuwen and N. Rummel, “Comparing teachers’ use of mirroring and advising dashboards,”
in ACM International Conference Proceeding Series, LAK’20, 2020, pp. 26–34.
[31] D. Etinger, “Discovering and mapping LMS course usage patterns to learning outcomes,” in IHSI
2020, 2020, vol. 1131 AISC, pp. 486–491.
[32] A. Bogarín, R. Cerezo, and C. Romero, “Discovering learning processes using inductive miner: A
case study with learning management systems (LMSs),” Psicothema, vol. 30, no. 3, pp. 322–329,
2018.
[33] N. Ahmad Uzir et al., “Discovering Time Management Strategies in Learning Processes Using
Process Mining Techniques,” in Transforming Learning with Meaningful Technologies. EC-TEL
2019. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence
and Lecture Notes in Bioinformatics), 2019, vol. 11722 LNCS, no. September, pp. 555–569.
[34] E. M. H. Real, E. Pinheiro Pimentel, L. V. De Oliveira, J. Cristina Braga, and I. Stiubiener,
“Educational Process Mining for Verifying Student Learning Paths in an Introductory Programming
Course,” in Proceedings - Frontiers in Education Conference, FIE, 2020, vol. 2020-Octob.
[35] A. Bogarín, C. Romero, R. Cerezo, and M. Sánchez-Santillán, “Clustering for improving Educational
process mining,” in ACM International Conference Proceeding Series, 2014, pp. 11–15.
[36] Y. Wang, T. Li, C. Geng, and Y. Wang, “Evaluating Student Learning Effect Based on Process
Mining,” in Applied Informatics. ICAI 2019. Communications in Computer and Information Science,
2019, pp. 59–72.
[37] Y. Wang, T. Li, C. Geng, and Y. Wang, “Recognizing patterns of student’s modeling behaviour
patterns via process mining,” Smart Learn. Environ., vol. 6, no. 1, 2019.
[38] C. Schoor and M. Bannert, “Exploring regulatory processes during a computer-supported
collaborative learning task using process mining,” Comput. Human Behav., vol. 28, no. 4, pp. 1321–
1331, 2012.
[39] K. Slaninová, J. Martinovič, P. Dráždilová, and V. Snašel, “From Moodle Log File to the Students
Network,” in International Joint Conference SOCO’13-CISIS’13-ICEUTE’13, 2014, vol. 239, pp.
641–650.
[40] A. R. C. Maita, M. Fantinato, S. M. Peres, L. H. Thom, and P. C. K. Hung, “Mining unstructured
processes: An exploratory study on a distance learning domain,” in Proceedings of the International
Joint Conference on Neural Networks, 2017, pp. 3240–3247.
[41] S. López‐Pernas, M. Saqr, and O. Viberg, “Putting it all together: Combining learning analytics
methods and data sources to understand students’ approaches to learning programming,” Sustain.,
vol. 13, no. 9, 2021.
[42] S. Shabaninejad, H. Khosravi, S. J. J. Leemans, S. Sadiq, and M. Indulska, “Recommending
insightful drill-downs based on learning processes for learning analytics dashboards,” in International
Conference on Artificial Intelligence in Education, 2020, vol. 12163 LNAI, no. June, pp. 486–499.
[43] R. Dolak, “Using Process Mining Techniques to Discover Student’s Activities, Navigation Paths, and
Behavior in LMS Moodle,” in Second International Conference, ICITL 2019, 2019, vol. 11937
LNCS, pp. 129–138.
[44] L. Juhaňák, J. Zounek, and L. Rohlíková, “Using process mining to analyze students’ quiz-taking
behavior patterns in a learning management system,” Comput. Human Behav., vol. 92, pp. 496–506,

99
2019.
[45] P. Reimann, J. Frerejean, and K. Thompson, “Using process mining to identify models of group
decision making in chat data,” in Proceedings of the 9th international conference on Computer
supported collaborative learning - Volume 1 (CSCL’09), 2009, pp. 98–107.
[46] A. van den Beemt, J. Buys, and W. van der Aalst, “Analysing structured learning behaviour in
Massive Open Online Courses (MOOCs): An approach based on process mining and clustering,” Int.
Rev. Res. Open Distance Learn., vol. 19, no. 5, pp. 38–60, 2018.
[47] F. Bernal, J. Maldonado-Mahauad, K. Villalba-Condori, M. Zúñiga-Prieto, J. Veintimilla-Reyes, and
M. Mejía, “Analyzing Students’ Behavior in a MOOC Course: A Process-Oriented Approach,” in
HCII 2020: HCI International 2020, 2020, vol. 12425 LNCS, pp. 307–325.
[48] G. Deeva, M. Snoeck, and J. De Weerdt, “Discovering the impact of students’ modeling behavior on
their final performance,” in The Practice of Enterprise Modeling. PoEM 2018. Lecture Notes in
Business Information Processing, 2018, vol. 335, pp. 335–350.
[49] K. Okoye and S. Hosseini, “Educational process intelligence: A process mining approach and model
analysis,” in IBICA 2019, 2021, vol. 1180 AISC, pp. 201–212.
[50] S. Rizvi, B. Rienties, J. Rogaten, and R. F. Kizilcec, “Investigating variation in learning processes in
a FutureLearn MOOC,” J. Comput. High. Educ., vol. 32, no. 1, pp. 162–181, 2020.
[51] J. Maldonado-Mahauad, M. Pérez-Sanagustín, R. F. Kizilcec, N. Morales, and J. Munoz-Gama,
“Mining theory-based patterns from Big data: Identifying self-regulated learning strategies in
Massive Open Online Courses,” Comput. Human Behav., vol. 80, pp. 179–196, 2018.
[52] S. Rizvi, B. Rienties, and J. Rogaten, “Temporal dynamics of MOOC learning trajectories,” ACM Int.
Conf. Proceeding Ser., no. October, 2018.
[53] S. S. Beheshitha, D. Gašević, and M. Hatala, “A process mining approach to linking the study of
aptitude and event facets of self-regulated learning,” ACM Int. Conf. Proceeding Ser., vol. 16-20-
Marc, pp. 265–269, 2015.
[54] L. Huang and S. P. Lajoie, “Process analysis of teachers’ self-regulated learning patterns in
technological pedagogical content knowledge development,” Comput. Educ., vol. 166, no. November
2020, p. 104169, 2021.
[55] K. Okoye, A. R. H. Tawil, U. Naeem, R. Bashroush, and E. Lamine, “A semantic rule-based
approach supported by process mining for personalised adaptive learning,” Procedia Comput. Sci.,
vol. 37, no. December, pp. 203–210, 2014.
[56] R. Rajendran, A. Munshi, M. Emara, and G. Biswas, “A temporal model of learner behaviors in
OELEs using process mining,” in ICCE 2018 - 26th International Conference on Computers in
Education, Main Conference Proceedings, 2018, no. December, pp. 276–285.
[57] H. Jeong, G. Biswas, J. Johnson, and L. Howard, “Analysis of productive learning behaviors in a
structured inquiry cycle using hidden Markov models,” Educ. Data Min. 2010 - 3rd Int. Conf. Educ.
Data Min., no. October, pp. 81–90, 2010.
[58] K. Engelmann and M. Bannert, “Analyzing temporal data for understanding the learning process
induced by metacognitive prompts,” Learn. Instr., vol. 72, no. October 2018, p. 101205, 2021.
[59] J. M. E. M. Van Der Werf, C. Van Schuppen, S. Brinkkemper, S. Jansen, P. Boon, and G. Van Der
Plas, “Architectural intelligence: A framework and application to e-learning,” in CEUR Workshop
Proceedings, 2017, vol. 1859, pp. 95–102.
[60] G. Sedrakyan, M. Snoeck, and J. De Weerdt, “Process mining analysis of conceptual modeling
behavior of novices - Empirical study using JMermaid modeling and experimental logging
environment,” Comput. Human Behav., vol. 41, pp. 486–503, 2014.
[61] T. Doleck, R. B. Basnet, E. G. Poitras, and S. P. Lajoie, “Mining learner–system interaction data:
implications for modeling learner behaviors and improving overlay models,” J. Comput. Educ., vol.
2, no. 4, pp. 421–447, 2015.
[62] G. Deeva and J. De Weerdt, “Understanding Automated Feedback in Learning Processes by Mining
Local Patterns,” in BPM 2018 International Workshops. Lecture Notes in Business Information
Processing, vol. 342, 2019, pp. 56–68.
[63] W. Poncin, A. Serebrenik, and M. Van Den Brand, “Mining student capstone projects with FRASR
and ProM,” in SPLASH’11 Compilation - Proceedings of OOPSLA’11, Onward! 2011, GPCE’11,
DLS’11, and SPLASH’11 Companion, 2011, pp. 87–95.
[64] A. Neyem, J. Diaz-Mosquera, J. Munoz-Gama, and J. Navon, “Understanding Student Interactions in
Capstone Courses to Improve Learning Experiences CSS Concepts • Social and professional topics
→ Project and people management • Software and its engineering → Software creation and

100
management General Terms,” in SIGCSE 2017 - Proceedings of the 48th ACM Technical Symposium
on Computer Science Education, 2017, pp. 423–428.
[65] J. Chen, Y. Zhang, J. Sun, Y. Chen, F. Lin, and Q. Jin, “Personalized micro-learning support based on
process mining,” Proc. - 2015 7th Int. Conf. Inf. Technol. Med. Educ. ITME 2015, pp. 511–515,
2016.
[66] M. Bannert, P. Reimann, and C. Sonnenberg, “Process mining techniques for analysing patterns and
strategies in students’ self-regulated learning,” Metacognition Learn., vol. 9, no. 2, pp. 161–185,
2014.
[67] C. Sonnenberg and M. Bannert, “Using Process Mining to examine the sustainability of instructional
support: How stable are the effects of metacognitive prompting on self-regulatory behavior?,”
Comput. Human Behav., vol. 96, pp. 259–272, 2019.
[68] S. Estupiñán and N. Szilas, “When did i lose them? using process mining to study user engagement in
interactive digital Narratives,” in Interactive Storytelling. 12th International Conference on
Interactive Digital Storytelling, ICIDS 2019, 2019, vol. 11869 LNCS, pp. 374–378.
[69] H. Nguyen, K. Y. Lim, L. L. Wu, C. Fischer, and M. Warschauer, “‘We’re looking good’: Social
exchange and regulation temporality in collaborative design,” Learn. Instr., vol. 74, no. October
2020, p. 101443, 2021.
[70] W. Matcha, D. Gašević, N. A. Uzir, J. Jovanović, and A. Pardo, “Analytics of learning strategies:
Associations with academic performance and feedback,” ACM Int. Conf. Proceeding Ser., pp. 461–
470, 2019.
[71] J. Saint, D. Gaševic, W. Matcha, N. A. A. Uzir, and A. Pardo, “Combining analytic methods to
unlock sequential and temporal patterns of self-regulated learning,” in Proceedings of the 10th
International Conference on Learning Analytics & Knowledge (LAK’20), 2020, pp. 402–411.
[72] R. Martinez-Maldonado, K. Yacef, and J. Kay, “Data mining in the classroom: Discovering groups’
strategies at a multi-tabletop environment,” in Proceedings of the 6th International Conference on
Educational Data Mining, EDM 2013, 2013, no. July 2015.
[73] J. Saint, D. Gašević, and A. Pardo, “Detecting Learning Strategies Through Process Mining,” in
Lifelong Technology-Enhanced Learning. EC-TEL 2018. Lecture Notes in Computer Science, 2018,
vol. 11082 LNCS, pp. 385–398.
[74] M. C. Rodriguez, M. L. Nistal, F. A. M. Fonte, M. L. Penin, and M. M. Molina, “Exploring the
application of process mining to support self-regulated learning: An initial analysis with video
lectures,” in IEEE Global Engineering Education Conference, EDUCON, 2018, vol. 2018-April, no.
August 2019, pp. 1766–1774.
[75] A. Bolt, M. de Leoni, W. M. P. van der Aalst, and P. Gorissen, “Business process reporting using
process mining, analytic workflows and process cubes: A case study in education,” in SIMPDA 2015,
2017, vol. 244, pp. 28–53.
[76] W. M. van der Aalst, S. Guo, and P. Gorissen, “Comparative Process Mining in Education: An
Approach Based on Process Cubes,” in International Symposium on Data-Driven Process Discovery
and Analysis, 2013, vol. 203, pp. 110–134.
[77] S. Heirweg, M. De Smul, E. Merchie, and G. Devos, Mine the process : investigating the cyclical
nature of upper primary school students ’ self ‑ regulated learning, no. 48. Springer Netherlands,
2020.
[78] R. Lira et al., “Process-oriented feedback through process mining for surgical procedures in medical
training: The ultrasound-guided central venous catheter placement case †,” Int. J. Environ. Res.
Public Health, vol. 16, no. 11, pp. 1–21, 2019.
[79] Y. Dahari, A. Gal, and A. Senderovich, “EDU-ProM: ProM for the classroom,” in CEUR Workshop
Proceedings, 2017, vol. 1920.
[80] F. Babič, J. Paralič, P. Bednár, and M. Raček, “Analytical framework for mirroring and reflection of
user activities in E-learning environment,” Adv. Intell. Soft Comput., vol. 80, no. September, pp. 287–
296, 2010.
[81] C. M. Chen and W. F. Wang, “Mining Effective Learning Behaviors in a Web-Based Inquiry Science
Environment,” J. Sci. Educ. Technol., vol. 29, no. 4, pp. 519–535, 2020.
[82] W. Premchaiswadi and P. Porouhan, “Process modeling and decision mining in a collaborative
distance learning environment,” Decis. Anal., vol. 2, no. 1, 2015.
[83] W. Peeters, M. Saqr, and O. Viberg, “Applying learning analytics to map students’ self-regulated
learning tactics in an academic writing course,” ICCE 2020 - 28th Int. Conf. Comput. Educ. Proc.,
vol. 1, no. September, pp. 245–254, 2020.

101
[84] Agusriandi, I. S. Sitanggang, and S. H. Wijaya, “Student Performance Based on Activity Log on
Social Network and e-Learning,” Int. J. Adv. Sci. Eng. Inf. Technol., vol. 10, no. 6, pp. 2276–2281,
2020.
[85] M. Mittal and A. Sureka, “Process mining software repositories from student projects in an
undergraduate software engineering course,” in 36th International Conference on Software
Engineering, ICSE Companion 2014 - Proceedings, 2014, pp. 344–353.
[86] J. A. Caballero-Hernández, A. Balderas, M. Palomo-Duarte, P. Delatorre, A. J. Reinoso, and J. M.
Dodero, “Teamwork assessment in collaborative projects through process mining techniques,” Int. J.
Eng. Educ., vol. 36, no. 1B, pp. 470–482, 2020.
[87] J. A. C. Hernández, M. P. Duarte, and J. M. Dodero, “An architecture for skill assessment in serious
games based on Event Sequence Analysis,” ACM Int. Conf. Proceeding Ser., vol. Part F1322, 2017.
[88] J. A. Ruipérez-Valiente, L. Rosenheck, and Y. J. Kim, “What Does Exploration Look Like? Painting
a Picture of Learning Pathways Using Learning Analytics,” in Game-Based Assessment Revisited, no.
November, 2019, pp. 281–300.
[89] T. Schaedler Uhlmann, E. Alves Portela Santos, and L. A. Mendes, “Process Mining Applied to
Player Interaction and Decision Taking Analysis in Educational Remote Games,” in Smart Industry
& Smart Education. REV 2018, 2019, vol. 47, pp. 425–434.
[90] S. A. Priyambada, E. R. Mahendrawathi, and B. N. Yahya, “Curriculum Assessment of Higher
Educational Institution Using Aggregate Profile Clustering,” in 4th Information Systems International
Conference 2017, ISICO 2017, 2017, vol. 124, pp. 264–273.
[91] L. Y. Bendatu and B. N. Yahya, “Sequence Matching Analysis for Curriculum Development,” J. Tek.
Ind., vol. 17, no. 1, pp. 47–52, 2015.
[92] N. Trčka and M. Pechenizkiy, “From local patterns to global models: Towards domain driven
educational process mining,” in ISDA 2009 - 9th International Conference on Intelligent Systems
Design and Applications, 2009, pp. 1114–1119.
[93] M. Cameranesi, C. Diamantini, L. Genga, and D. Potena, “Students’ Careers Analysis: a Process
Mining Approach,” in Proceedings of the 7th International Conference on Web Intelligence, Mining
and Semantics (WIMS ’17), 2017, pp. 1–7.
[94] J. P. Salazar-Fernandez, M. Sepúlveda, J. Munoz-Gama, and M. Nussbaum, “Curricular analytics to
characterize educational trajectories in high-failure rate courses that lead to late dropout,” Appl. Sci.,
vol. 11, no. 4, pp. 1–15, 2021.
[95] J. P. Salazar-Fernandez, M. Sepúlveda, and J. Munoz-Gama, “Describing educational trajectories of
engineering students in individual high-failure rate courses that lead to late dropout,” in CEUR
Workshop Proceedings, 2019, vol. 2425, pp. 39–48.
[96] J. A. Caballero-Hernandez, J. M. Dodero, I. Ruiz-Rube, M. Palomo-Duarte, J. F. Argudo, and J. J.
Dominguez-Jimenez, “Discovering bottlenecks in a computer science degree through process mining
techniques,” in SIIE 2018 - 2018 International Symposium on Computers in Education, Proceedings,
2018, no. 1, pp. 1–6.
[97] V. Aisa, A. P. Kurniati, and A. W. Yanuar Firdaus, “Evaluation of the online assessment test using
process mining (Case Study: Intensive English Center),” in 2015 3rd International Conference on
Information and Communication Technology, ICoICT 2015, 2015, pp. 472–477.
[98] A. Piotrkowicz, V. Dimitrova, and T. E. Roberts, “Temporal Analytics of Workplace-Based
Assessment Data to Support Self-regulated Learning,” in EC-TEL 2018: 13th European Conference
on Technology Enhanced Learning. Lecture Notes in Computer Science, 2018, vol. 11082 LNCS, pp.
570–574.
[99] A. H. Cairns et al., “Using semantic lifting for improving educational process models discovery and
analysis,” in CEUR Workshop Proceedings, 2014, vol. 1293, pp. 150–161.
[100]A. Hicheur Cairns, B. Gueni, H. Hafdi, C. Joubert, and N. Khelifa, “Towards a distributed
computation platform tailored for educational process discovery and analysis,” in International
Conference on Protocol Engineering, ICPE 2015 and International Conference on New Technologies
of Distributed Systems, NTDS 2015 - Proceedings, 2015, pp. 1–8.
[101]J. Munoz-Gama, V. Gálvez, R. de la F. Sanhueza, M. Sepulveda, and R. S. Fuentes, “Interactive
Process Mining for Medical Training,” in Interactive Process Mining in Healthcare, C. Fernandez-
Llatas, Ed. Springer, Cham, 2021, pp. 233–242.
[102]A. Sypsas and D. Kalles, “Computing Similarities Between Virtual Laboratory Experiments Models
using Petri Nets,” in 20th International Conference on Modelling and Applied Simulation, MAS 2021,
2021, pp. 29–37.

102
[103]D. Riofrío-Luzcando and J. Ramírez, “Predictive student action model for procedural training in 3D
virtual environments,” in E-Learning, E-Education, and Online Training. Second International
Conference, eLEOT 2015, 2016, pp. 37–62.
[104]J. C. Vidal, M. Lama, B. Vázquez, and M. Mucientes, “Reconstructing IMS LD units of learning
from event logs,” in Proceedings of the 9th European Conference on Open Learning and Teaching in
Educational Communities, 2014, vol. 8719 LNCS, pp. 345–358.
[105]J. C. Vidal, B. Vázquez-Barreiros, M. Lama, and M. Mucientes, “Recompiling learning processes
from event logs,” Knowledge-Based Syst., vol. 100, no. C (May 2016), pp. 160–174, 2016.
[106]L. Castillo, “A virtual laboratory for multiagent systems: Joining efficacy, learning analytics and
student satisfaction,” in 2016 International Symposium on Computers in Education, SIIE 2016:
Learning Analytics Technologies, 2016, pp. 1–6.
[107]B. M. Toussaint and V. Luengo, “Mining surgery phase-related sequential rules from vertebroplasty
simulations traces,” in 15th International Conference on Artificial Intelligence in Medicine, 2015,
vol. 9105, pp. 35–46.
[108]B. Fernández-Gallego, M. Lama, J. C. Vidal, and M. Mucientes, “Learning analytics framework for
educational virtual worlds,” in International Conference on Virtual and Augmented Reality in
Education, 2013, vol. 25, pp. 443–447.
[109]R. Dijkman and A. Wilbik, “Linguistic summarization of event logs – A practical approach,” Inf.
Syst., vol. 67, pp. 114–125, 2017.
[110]W. Van der Aalst, Process mining: Data science in action, Second Edi. 2016.
AUTHORS
Athanasios Sypsas is a PhD candidate at Hellenic Open University, School of
Science and Technology, Greece. His main scientific interests are Simulation,
Artificial Intelligence, Distance Learning and Programming.
Dimitris Kalles is a Professor on “Artificial Intelligence – Applications” with the
Hellenic Open University (HOU) and Director of the undergraduate study programme
“Computer Science”. His main research interests are in Artificial Intelligence and
Machine Learning and his work also has a strong focus in Educational Technology and
Software Engineering. He has published over 100 papers in scientific journals and
conference proceedings and his work has received more than 1000 citations.

REVIEWING PROCESS MINING APPLICATIONS AND TECHNIQUES IN EDUCATION

More Related Content

What's hot (13)

Similar to REVIEWING PROCESS MINING APPLICATIONS AND TECHNIQUES IN EDUCATION (20)

Recently uploaded (20)

REVIEWING PROCESS MINING APPLICATIONS AND TECHNIQUES IN EDUCATION