Interpretation of Patients’ Location Data to Support the Application of
Process Mining Notations
Sina Namaki Araghi, Franck Fontanili, Elyes Lamine, Nicolas Salatge and Frederick Benaben
IMT Mines Albi, France
Keywords:
Model-driven Engineering, Process Mining, Indoor Localization Systems (ILS), Patient Pathways.
Abstract:
The application of indoor localization and process mining emerges as an intriguing tool for the researchers
to address the structural issues related to the patient pathways inside healthcare organizations. However,
there is a major gap in the literature. This is related to the lack of enough attention to the interpretation of
location data. Therefore, as a contribution, this article presents the DIAG meta-model and relevant location
data interpretation rules. This model-driven approach has been realized in the context of the R-IOSUITE
application and it supports the further analyses by the process mining methods.
1 INTRODUCTION
The so-called ”no-shows” or ”missed appointments“
concept is a costly defect for patients and healthcare
organization. According to the article published in
(Anisi et al., 2018), the ratio of the missed appoint-
ments is about 34 % only in the USA, and each missed
appointment costs about 100 $. One of the major
causes behind this issue is the structural problems
of the healthcare processes.
Recently, the application of Indoor Localiza-
tion Systems (ILS) and Process Mining is pro-
posed in several research works to address such is-
sues and to improve the outcome of patient pathways
(Kamel Boulos and Berry, 2012; Martin, 2019; Rojas
et al., 2017b; Martinez-Millana et al., 2019).
In line with these applications, this research work
highlights a major concern in this area. According to
the analysis of the literature provided in this article,
most of research works focused on the ”knowledge-
extraction” approaches by using location data of the
tagged objects. On the other hand, the interpretation
of location data or as called here the ”sense-making
procedures” are completely ignored.
Therefore, this research work aims to propose a
model-driven engineering approach to tackle this is-
sue. As a result, the contribution of this article is ori-
ented towards the DIAG meta-model which supports
the interpretation of the location data that can be
performed by two groups of interpretation rules.
The used methodology in this article is shown in
figure 1. This methodology which has been devel-
oped previously in (Namaki Araghi et al., 2018b; Na-
maki Araghi et al., 2019) consists of several functions
to extract meaningful information by using indoor lo-
calization systems in association with process mining.
These functions are: configuring the environment and
systems, Location data gathering , Location data in-
terpreting, Business process modeling, and Business
process analyzing and diagnosing. The focus of this
paper is mainly on the first and third functions. Within
these functions, the contributions of this article pre-
pare a solid foundation prior to applying process min-
ing notations.
Figure 1: The used approach in this research work.
Consequently, the next section of this article
presents the analysis of the literature. The third sec-
472
Araghi, S., Fontanili, F., Lamine, E., Salatge, N. and Benaben, F.
Interpretation of Patients’ Location Data to Support the Application of Process Mining Notations.
DOI: 10.5220/0008971104720481
In Proceedings of the 13th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2020) - Volume 5: HEALTHINF, pages 472-481
ISBN: 978-989-758-398-8; ISSN: 2184-4305
Copyright
c
2022 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
Figure 2: The used approach for analyzing research works related to applying localization techniques for a better organization
of processes and activities in different industries.
tion focuses in presenting the DIAG meta-model and
the location data interpretation rules. The fourth sec-
tion presents a series of experimental results. Finally,
the last section concludes the article.
2 BACKGROUND
This section presents a systematic literature and tech-
nological review in a five-year (2015- June 2019) pe-
riod. The objective for choosing this period was to
recognize the recent trends in the field of applying lo-
cation data for data and process mining. The main
focus here was to analyze the existing works that ad-
dressed the trajectory pattern mining regardless of a
specific industry. These works were selected due to
the nearness of the presented methods to the applied
algorithms for discovery of the end-to-end process-
like patterns.
Accordingly, Google scholar and Web Of Science
were the two search engines used for this purpose.
Several keywords were considered such as process
mining, location data, indoor location/localization
systems. As a result, 42 research works were
recorded. We tried to ensure about the proximity of
the obtained works to the purposes of this research
work. Therefore, by skimming the abstract, conclu-
sion and introduction of these works, 31 out of 42
were selected for final literature analysis. Table 1 pro-
vides an overview of the cited works. Figure 2 shows
the approach for analyzing the literature related to the
works that have applied localization techniques.
As mentioned previously, the objective was to
capture an overall image of the recent trends for us-
ing location data in different industries. Notably, two
criteria were used; (i) works related to the interpreta-
tion of location data, and (ii) works which addressed
the knowledge-extraction based on location data.
In some cases researchers worked on several of
the mentioned criteria. For instance the work in
(Senderovich et al., 2016; Wan et al., 2017; Muza-
mmal et al., 2018), it addresses both mining process-
like patterns and how to interpret the location data.
According to the presented analysis in figure 3, the
recent trend is mainly on the extraction of knowl-
edge from location data. On the other hand, despite
its importance, the interpretation and preparation
of location data were almost completely neglected.
The following will present briefly some of the sig-
nificant information extracted from these works.
2.1 Preparation and Interpretation of
Location Data
Despite the lack of attention for this notion, several
robust works proposed different methods to interpret
the location data prior to the knowledge discovery.
In (Senderovich et al., 2016) the authors used a so-
called interaction mining to transform the real-time
location data into standard event logs for enabling
process mining activities. The transformation of event
streams was according to the defined notion of inter-
action. This interaction provides a knowledge layer
that links the raw sensor data and process instances.
In (Muzammal et al., 2018), the authors have
used the uncertain sensor data and tried to transform
these data into probabilistic trajectory data using pre-
processing routines. Then, they have used a dynamic
agorithmic approach to extract the preferable trajec-
tories. Their approach consists of 3 main steps: data
cleaning , data compression and finally trajectory
mining. The trajectory creation step which generates
togetherness pattern, common path pattern and group
patterns and cyclic patterns.
The work in (Wan et al., 2017) presents a frame-
work for deriving information about people continu-
ous activities from their individual GPS data. The au-
thors have proposed a framework which includes two
major techniques for processing GPS data: (i) a fuzzy
classification for differentiating the activity patterns .
(ii) a scale- adaptive method for refining the location
of activities within outdoor and indoor environments.
Interpretation of Patients’ Location Data to Support the Application of Process Mining Notations
473
Table 1: Cited research works that addressed the knowledge extraction from location data and those which addressed inter-
pretation of these data.
Cited research works
location event log preparation
and interpretation
(Senderovich et al., 2016; Muzammal et al., 2018; Wan et al., 2017;
Namaki Araghi et al., 2018a)
Use of location data for extraction
of knowledge
(Hwang and Jang, 2017; Sztyler et al., 2016; Ertek et al., 2017; Mazim-
paka and Timpf, 2016), (Rojas et al., 2017a; Zheng, 2015a; Tanuja and
Govindarajulu, 2017; Ramos et al., 2017; Garaeva et al., 2017; Bao and
Wang, 2017), (Zhenjiang et al., 2017; Aryal and Sujing Wang, 2017;
Tanuja and Govindarajulu, 2016; Feng Ling et al., 2016), (Lamr and
Skrbek, 2016; Blank et al., 2016; Fernandez-Llatas et al., 2015; Zheng,
2015b), (Liao et al., 2015; Tang et al., 2015; Miclo et al., 2015; Jin
et al., 2015; Martinez-Millana et al., 2019), (Dogan et al., 2019; Na-
maki Araghi et al., 2019; Namaki Araghi et al., 2018a; Namaki Araghi
et al., 2018b; Araghi et al., 2018)
Figure 3: Analysis of the literature of process mining, indoor localization systems, and data mining relevant to the approaches
in which the location data were used.
The clustering techniques used in their work helped to
classify which activities are occurring. It helped also
to detect the outliers and indicated in which positions
an activity is being executed.
The next section will highlight how the review of
these previous works gave the direction for the pro-
posed solutions in this article.
2.2 A Decision based on the Literature
The processing of location data is not only a technical
challenge, but it can be seen as a scientific challenge
as well. Due to the existing high ambiguity in location
data, it is not feasible to extract proper information
from event logs. This ambiguity can be caused by the
way these data are being registered.
These data are recorded by coordination. They
are not representing activities and they only illustrate
the positions of tags. Therefore, researchers need
to address how one can interpret this type of data.
Simply put, the sense-making procedure is miss-
ing in the literature. Some researchers insisted on
this important gap, and indicated the need to pro-
vide an abstraction of what exists in the location data
(Senderovich et al., 2016).
In line with this issue, the next section will focus
on presenting the DIAG meta-model and the location
data interpretation rules.
3 METHOD
Assume that a location event log similar to the one in
figure 4 is given for further analyses. By focusing on
such event log, one could observe that the given data
is vague. This means it is not clear which recorded
information belongs to the patients, which belongs to
the healthcare staff, and what is the corresponding ac-
tivity. In that, there exist many other blurred informa-
tion that should be detected. In fact, there is a con-
siderable gap to link the actual concepts in patient—
pathways—processes with the registered data in the
location event logs. Figure 4 embodies this gap be-
HEALTHINF 2020 - 13th International Conference on Health Informatics
474
Figure 4: Illustrating the need to provide a sense-making procedure for the recorded information in location event logs.
tween raw data and the actual concepts in patient pro-
cesses.
3.1 DIAG Meta-model
To fill this gap, the DIAG meta-model is presented in
figure 5. This meta-model consists of seven different
packages. Each package contains several classes that
help to create a model of the link between the actual
concepts in the environment and the raw data in loca-
tion event logs. As shown in figure 5, (i) Healthcare
resources package includes the main resources used
in patient processes that can be detected and moni-
tored by location tags. (ii) The Organization package
which represents the main classes that are linked to
the healthcare organization. (iii) Location event logs
package provides a model of all the information that
exist within a location event log and their relationship
with other classes. (iv) The Objective package con-
sists of necessary classes to perform business process
analyses on top of the registered information. (v) The
Process package models the concepts that are consid-
ered to represent a patient pathways as a business pro-
cess model. (vi) The Function package is addressing
the different classes that lead to the execution of ac-
tivities in processes. (vii) The Healthcare functions
package addresses different types of actions that can
take place during the run-time of patient pathways.
(viii) The Context package helps to provide a base
for further business process diagnosing actions.
The DIAG meta-model has been realized and de-
veloped thanks to a web-based application known as
R IOSUITE. This multidisciplinary framework in-
cludes multiple modules (R-IODA, R-IOSEMIT, R-
IOGA and etc.) to obtain the location event logs and
extract the required information. https://research-gi.
mines-albi.fr/display/RIOSUITE/R-IOSuite+Home
Evidently, the detailed presentation of the pack-
ages and their included classes are out of the limit of
this paper. However, some examples illustrate the ap-
plicability of this meta-model. Figure 6 shows how
the user can model engaged healthcare resources in
patient pathways by the application. As shown within
figure 6, a hospital is addressed by the Organiza-
tion package and it can have different Healthcare re-
sources such as healthcare staff and other resources
like the hospital zones. Modeling these concepts
within the application provides a-priori knowledge for
interpreting the information within the location event
logs.
3.2 Location Data Interpretation Rules
After gathering the primary location event logs, they
will be uploaded to each defined ”location tag“ class
in the ”resource“ tab of the application (c.f. figure 6).
As shown in figure 1, it is necessary to interpret the
location event logs in order to be used for the process
modeling function.
In line with this objective, two groups of location
interpretation rules are defined.
• First group: This set is constructed by three rules.
(i) Add start-event, (ii) Add end-event, and (iii)
Add task-events. They are in charge of trans-
forming the very first visualization of data as
”nodes“ with out any semantics. These graphs
will be sent to a graph database management sys-
tem. For example, outputs of this step will be se-
ries of nodes labeled as ”Activity.Inside.Zone # “.
• Second group: After detecting the nodes, the sec-
ond group of rules will start retracting the mod-
eled a-priori knowledge within the application.
These rules are: (i) Add knowledge onto start-
Interpretation of Patients’ Location Data to Support the Application of Process Mining Notations
475
Figure 5: The DIAG meta-model.
event, (ii) Add knowledge to end-event, and (iii)
Add knowledge to task-event. As mentioned
this action is possible by the modeled information
within the ”resource“ tab of the application.
Since the presentation of all the rules is much broader
than the limit of this paper, we only present one of the
most critical ones which is the Add task-events.
3.2.1 Add Task-event
The presented algorithm in this section (c.f. algorithm
1) describes how to discover the activities from the lo-
cation event logs. Assuming human functions are ex-
ecuting the process activities within defined zones of
the facility, one can infer that an activity (a task) starts
when a process actor enters a certain zone and it fin-
ishes when the process actor leaves that zone. Algo-
rithm 1 expresses in more detail the rule for detecting
such activities.
The algorithm is looking for the events that are re-
lated to the zones (measurement.update.zone). How-
ever, here the algorithm is seeking zones that are
within the general process zones or zones with the id
greater than ‘0’ (If the z
j
.id > 0).
After identifying the zone in z
1
event we look for
the next event that share the same zone id (z
j
.id), and
additionally, we ensure that the second event also is
related to the same patient (for example: z
3
.eui ==
z
6
.eui). This helps to ensure that the activity has been
finished and the algorithm is not considering an in-
complete data in the event log. This is an important
characteristic to provide a refined set of data sets for
avoiding the incompleteness notion in event logs.
HEALTHINF 2020 - 13th International Conference on Health Informatics
476
Algorithm 1: Add-task-event.
1: procedure ADDTASKEVENT Discovering the Human Task
2: Input < EventLog >= Structure
3: < Name >: string;
4: < Id >: integer;
5: < Tags >: integer;
6: < Topic >: string;
7:
8: Select(EventLog, Topic == “measurement.update.zone”)
9:
10: < zone event >= Structure;
11: < Topic >= string ;
12: < EventID >= integer;
13: j=1,2,3,...,m
14: < eui >= integer;
15: < DateTime >= string;
16: < InZone >= boolean;
17: < Zone.ID >= integer; i= 0,1,2,3,...,n
18: for zone event do
19: if (z
1
= zone event [z
1
.topic== ‘measurement.update.zone, AND z
1
.id > 0 ] AND (z
2
= zone event[z
2
.topic == ‘measurement.update.zone
0
AND z
2
.id > 0 AND z
1
.id==z
2
.id AND z
1
.eui == z
2
.eui]))
20: then
21: Select “z
1
.event id” as ‘HumanTask’ ;
22: Define ’name’ (“z
1
.event id” as “Activity.Inside.Zone”);
23: Register ’z
1
.event. propoerties’ as ‘HumanTask. propoerties’;
24: end if
25: if ((z
1
.id == z
2
.id) AND( z
1
.eui==z
2
.eui )AND (z
1
.zone.attribute in zone==‘TRUE
0
) AND (z
2
.zone.attribute in zone==‘FALSE
0
) AND
(z
1
.date time, z
2
.date time ))
26:
27: OR In case of existing noises
28:
29: (z
2
.zone.attribute in zone == ‘TRUE‘) AND z
1
.zone.attribute in zone==‘FALSE‘) AND (z
2
.date time, z
1
.date time)))
30: then
31: Register z
j
.date time;
32: end if
33: end for
34: end procedure
After extracting the starting point of an activity,
the algorithm randomly defines a node and name it as
”Activity.Inside.Zone #” and all the defined proper-
ties of the node such as time will be recorded. Note
that a process actor may go several times into a zone
and goes out during his/her process. In order to avoid
misinterpreting the activity sequences, the algorithm
uses the date time concept (at line 14 of the algorithm
1). At the final step, duration of the activity is regis-
tered thanks to the extracted start and end-event.
Later on, these nodes will be retracted by the
”data interpreter” in order to add more knowledge on
them. It is worth noting that, the a-priori knowledge
is provided thanks to the implementation of the DIAG
meta-model (c.f. 5).
4 RESULT
To evaluate the applicability of this approach, an ex-
periment was conducted within a hospital living-lab
in south of France. During this study, 30 patients
were monitored. To begin with this experiment, the
resources and other involved elements with the pa-
tient pathways were modeled in the R-IOSUITE ap-
plication. This action is similar to what has been
shown previously in figure 6. Next, each patient re-
ceived a location tag. Consequently, several location
event logs were registered by the localization system.
These data were uploaded in the application to be in-
terpreted.
An example of the interpretation procedure is
shown in figure 7. In figure 7 at the left side, a map
illustrates the movement of the patient who is wearing
the tag 3. On the right side the graph represents how
the events are seen. According to this graph each time
Interpretation of Patients’ Location Data to Support the Application of Process Mining Notations
477
Figure 6: An screen-shot of the R.IO-DIAG application which illustrate the modeling of the healthcare resources in the
application.
Figure 7: An illustration of the interpretation procedure.
a patient enters a zone the graph will reach the upper
threshold, and by touching the lower threshold, it can
be inferred that the patient left the zone.
After finishing the interpretation of the location
event logs, the application discovers the patient path-
ways by using the heuristic miner algorithm. This
well-known process discovery algorithm previously
has been presented in the work of (Weijters and
Ribeiro, 2011; Weijters et al., 2006; De Cnudde et al.,
2014).
Figure 8 shows the result of the process modeling
function. The patient pathways in this research work
were modeled by the OPC (Operation Process Chart)
modeling language (Namaki Araghi et al., 2018a).
Thanks to the development of the meta-model, the
application is able to detect what are the different
types of activities that patient are doing within their
process. These activity types are defined within the
Healthcare function package. This supports the fur-
ther process mining analyses; such as, which class of
HEALTHINF 2020 - 13th International Conference on Health Informatics
478
Figure 8: The process model extracted from the location event logs.
activities is leading to deviation in patient pathways.
Diagnosing such deviations could unfold the further
improvement plans.
5 CONCLUSIONS
The association of localization technologies and pro-
cess mining in healthcare is limited by two main cat-
egories of different issues. The first category is the
social challenges regarding the use of location data
of patients or staff. There are multiple aspects that
should be addressed such as the ethical issues. The
second category is the scientific and technical chal-
lenges.
Concerning this category, there exists a major is-
sue which is the inadequate attention in the litera-
ture to the ”interpretation of location data for pro-
cess mining“. As a result, this research work pro-
posed the DIAG meta-model which can provide a
proper core to interpret the location data and support
the sense-making procedure. Such model-driven ap-
proaches were encouraged previously for enriching
the raw data in data science projects (Benaben et al.,
2019).
To best of our knowledge, such a model-driven
approach has not been used previously for apply-
ing process mining and indoor localization technolo-
gies to monitor patient pathways. Indeed, due to the
complexity of the healthcare organizations, this meta-
model and the corresponding interpretation rules may
not be a global answer for different hospitals. How-
ever, it can reach a sufficient maturity level by receiv-
ing more practical experiences.
REFERENCES
Anisi, S., Zarei, E., Sabzi, M., Chehrazi, M., et al. (2018).
Missed appointments: Factors contributing to patient
no-show in outpatient hospital clinics in tehran, iran.
Shiraz E-Medical Journal, 19(8).
Araghi, S. N., Fontanili, F., Lamine, E., Tancerel, L., and
Benaben, F. (2018). Monitoring and analyzing pa-
tients’ pathways by the application of Process Mining,
SPC, and I-RTLS. IFAC-PapersOnLine, 51(11):980–
985.
Aryal, A. M. and Sujing Wang (2017). Discovery of pat-
terns in spatio-temporal data using clustering tech-
niques. In 2017 2nd International Conference on Im-
age, Vision and Computing (ICIVC), pages 990–995.
Bao, X. and Wang, L. (2017). Discovering Interesting Co-
location Patterns Interactively Using Ontologies. In
Bao, Z., Trajcevski, G., Chang, L., and Hua, W., ed-
itors, Database Systems for Advanced Applications,
pages 75–89, Cham. Springer International Publish-
ing.
Benaben, F., Li, J., Koura, I., Montreuil, B., Lauras, M.,
Mu, W., and Gou, J. (2019). A Tentative Framework
for Risk and Opportunity Detection in A Collaborative
Environment Based on Data Interpretation.
Blank, P., Maurer, M., Siebenhofer, M., Rogge-Solti, A.,
and Schonig, S. (2016). Location-Aware Path Align-
ment in Process Mining. In 2016 IEEE 20th In-
ternational Enterprise Distributed Object Computing
Workshop (EDOCW), pages 1–8.
De Cnudde, S., Claes, J., and Poels, G. (2014). Improv-
ing the Quality of the Heuristics Miner in ProM 6.2.
Expert Syst. Appl., 41(17):7678–7690.
Dogan, O., Bayo-Monton, J.-L., Fernandez-Llatas, C., and
Oztaysi, B. (2019). Analyzing of Gender Behaviors
from Paths Using Process Mining: A Shopping Mall
Application. Sensors (Basel, Switzerland), 19(3).
Ertek, G., Chi, X., and Zhang, A. N. (2017). A Frame-
work for Mining RFID Data From Schedule-Based
Systems. IEEE Transactions on Systems, Man, and
Cybernetics: Systems, 47(11):2967–2984.
Feng Ling, Tianyue Sun, Xinning Zhu, Qingqing Chen, Xi-
aosheng Tang, and Xin Ke (2016). Mining travel be-
haviors of tourists with mobile phone data: A case
Interpretation of Patients’ Location Data to Support the Application of Process Mining Notations
479
study in Hainan. In 2016 2nd IEEE International Con-
ference on Computer and Communications (ICCC),
pages 1524–1529.
Fernandez-Llatas, C., Lizondo, A., Monton, E., Benedi, J.-
M., and Traver, V. (2015). Process mining methodol-
ogy for health process tracking using real-time indoor
location systems. Sensors, 15(12):29821–29840.
Garaeva, A., Makhmutova, F., Anikin, I., and Sattler, K.
(2017). A framework for co-location patterns min-
ing in big spatial data. In 2017 XX IEEE Interna-
tional Conference on Soft Computing and Measure-
ments (SCM), pages 477–480.
Hwang, I. and Jang, Y. J. (2017). Process Mining to Dis-
cover Shoppers’ Pathways at a Fashion Retail Store
Using a WiFi-Base Indoor Positioning System. IEEE
Transactions on Automation Science and Engineer-
ing, 14(4):1786–1792.
Jin, P., Du, J., Huang, C., Wan, S., and Yue, L. (2015). De-
tecting hotspots from trajectory data in indoor spaces.
In International Conference on Database Systems for
Advanced Applications, pages 209–225. Springer.
Kamel Boulos, M. N. and Berry, G. (2012). Real-time
locating systems (RTLS) in healthcare: a condensed
primer. International Journal of Health Geographics,
11:25.
Lamr, M. and Skrbek, J. (2016). Traffic Data and
Possibilities of Their Utilization for Safer Traf-
fic. Technical Univ Liberec, Faculty Economics.
WOS:000404420200007.
Liao, J., Wang, Z., Wan, L., Cao, Q. C., and Qi, H. (2015).
Smart Diary: A Smartphone-Based Framework for
Sensing, Inferring, and Logging Users’ Daily Life.
IEEE Sensors Journal, 15(5):2761–2773.
Martin, N. (2019). Using Indoor Location System Data
to Enhance the Quality of Healthcare Event Logs:
Opportunities and Challenges. In Daniel, F., Sheng,
Q. Z., and Motahari, H., editors, Business Process
Management Workshops, Lecture Notes in Business
Information Processing, pages 226–238. Springer In-
ternational Publishing.
Martinez-Millana, A., Lizondo, A., Gatta, R., Vera, S., Sal-
cedo, V. T., and Fernandez-Llatas, C. (2019). Process
Mining Dashboard in Operating Rooms: Analysis of
Staff Expectations with Analytic Hierarchy Process.
International Journal of Environmental Research and
Public Health, 16(2).
Mazimpaka, J. D. and Timpf, S. (2016). Trajectory data
mining: A review of methods and applications. Jour-
nal of Spatial Information Science, 2016(13):61–99.
Miclo, R., Fontanili, F., Marqu
`
es, G., Bomert, P., and
Lauras, M. (2015). RTLS-based Process Mining: To-
wards an automatic process diagnosis in healthcare. In
2015 IEEE International Conference on Automation
Science and Engineering (CASE), pages 1397–1402.
Muzammal, M., Gohar, M., Rahman, A. U., Qu, Q., Ah-
mad, A., and Jeon, G. (2018). Trajectory Mining Us-
ing Uncertain Sensor Data. IEEE Access, 6:4895–
4903.
Namaki Araghi, S., Fontanili, F., Lamine, E., Salatge, N.,
Lesbegueries, J., Pouyade, S. R., and Benaben, F.
(2019). Evaluating the Process Capability Ratio of
Patients’ Pathways by the Application of Process Min-
ing, SPC and RTLS. pages 302–309.
Namaki Araghi, S., Fontanili, F., Lamine, E., Salatge, N.,
Lesbegueries, J., Pouyade, S. R., Tancerel, L., and
Benaben, F. (2018a). A Conceptual Framework to
Support Discovering of Patients’ Pathways as Oper-
ational Process Charts. In 2018 IEEE/ACS 15th Inter-
national Conference on Computer Systems and Appli-
cations (AICCSA), pages 1–6.
Namaki Araghi, S., Fontanili., F., Lamine., E., Tancerel., L.,
and Benaben., F. (2018b). Applying process mining
and rtls for modeling, and analyzing patients’ path-
ways. In Proceedings of the 11th International Joint
Conference on Biomedical Engineering Systems and
Technologies - Volume 5: HEALTHINF,, pages 540–
547. INSTICC, SciTePress.
Ramos, J., C
´
esar, A., Neves, J., and Novais, P. (2017).
Adapting the User Path Through Trajectory Data Min-
ing. In De Paz, J. F., Juli
´
an, V., Villarrubia, G.,
Marreiros, G., and Novais, P., editors, Ambient In-
telligence– Software and Applications – 8th Inter-
national Symposium on Ambient Intelligence (ISAmI
2017), Advances in Intelligent Systems and Comput-
ing, pages 195–202. Springer International Publish-
ing.
Rojas, E., Fern
´
andez-Llatas, C., Traver, V., Munoz-Gama,
J., Sep
´
ulveda, M., Herskovic, V., and Capurro, D.
(2017a). Palia-er: Bringing question-driven process
mining closer to the emergency room. In BPM (De-
mos).
Rojas, E., Sep
´
ulveda, M., Munoz-Gama, J., Capurro,
D., Traver, V., and Fernandez-Llatas, C. (2017b).
Question-Driven Methodology for Analyzing Emer-
gency Room Processes Using Process Mining. Ap-
plied Sciences, 7(3):302.
Senderovich, A., Rogge-Solti, A., Gal, A., Mendling, J.,
and Mandelbaum, A. (2016). The ROAD from Sen-
sor Data to Process Instances via Interaction Mining.
In Nurcan, S., Soffer, P., Bajec, M., and Eder, J.,
editors, Advanced Information Systems Engineering,
pages 257–273, Cham. Springer International Pub-
lishing.
Sztyler, T., Carmona, J., V
¨
olker, J., and Stuckenschmidt,
H. (2016). Self-tracking reloaded: applying process
mining to personalized health care from labeled sen-
sor data.
Tang, L.-A., Yu, X., Gu, Q., Han, J., Jiang, G., Leung, A.,
and Porta, T. L. (2015). A Framework of Mining Tra-
jectories from Untrustworthy Data in Cyber-Physical
System. ACM Trans. Knowl. Discov. Data, 9(3):16:1–
16:35.
Tanuja, V. and Govindarajulu, P. (2016). Application of
trajectory data mining techniques in crm using move-
ment based community clustering. 16(11):20.
Tanuja, V. and Govindarajulu, P. (2017). A novel frame-
work for geo-clustering of user movements based on
trajectory data. 17(3):212.
Wan, N., Kan, G. L., and Wilson, G. (2017). Addressing
location uncertainties in GPS-based activity monitor-
ing: A methodological framework. Transactions in
GIS, 21(4):764–781.
Weijters, A., van der Aalst, W. M. P., and De Medeiros,
HEALTHINF 2020 - 13th International Conference on Health Informatics
480
A. A. (2006). Process mining with the heuristics
miner-algorithm. 166:1–34.
Weijters, A. J. M. M. and Ribeiro, J. T. S. (2011). Flexi-
ble Heuristics Miner (FHM). In 2011 IEEE Sympo-
sium on Computational Intelligence and Data Mining
(CIDM), pages 310–317.
Zheng, Y. (2015a). Trajectory Data Mining: An Overview.
ACM Trans. Intell. Syst. Technol., 6(3):29:1–29:41.
Zheng, Y. (2015b). Trajectory Data Mining: An Overview.
ACM Transaction on Intelligent Systems and Technol-
ogy.
Zhenjiang, D., Deng, J., Xiaohui, J., and Yongli, W. (2017).
RTMatch: Real-time location prediction based on tra-
jectory pattern matching. In Database Systems for
Advanced Applications - DASFAA 2017 International
Workshops: BDMS, BDQM, SeCoP, and DMMOOC,
Proceedings, pages 103–117. Springer Verlag.
Interpretation of Patients’ Location Data to Support the Application of Process Mining Notations
481
