An essential component for evaluating and comparing physical and cognitive capabilities between populations is the testing of various factors related to human performance. As a core part of sports science research, testing motor performance enables the analysis of the physical health of diferent demographic groups and makes them comparable. The Motor Research (MO|RE) data repository, developed at the Karlsruhe Institute of Technology, is an infrastructure for publishing and archiving research data in sports science, particularly in the field of motor performance research. In this paper, we present our vision for creating a knowledge graph from MO|RE data. With an ontology rooted in the Basic Formal Ontology, our approach centers on formally representing the interrelation of plan specifications, specific processes, and related measurements. Our goal is to transform how motor performance data are modeled and shared across studies, making it standardized and machineunderstandable. The idea presented here is developed within the Leibniz Science Campus “Digital Transformation of Research” (DiTraRe).
Research has undergone significant changes under the influence of ongoing digitalization; it has afected
how scientists conduct and share research, be it in the form of digital preprints [
Recent advances in artificial intelligence (AI) and the increasing prevalence of large language models
(LLMs) underscore the need for improvements beyond traditional approaches to data processing and
structuring. This enables machine-understandability, as current datasets are often too limited to support
sophisticated automated analysis with these technologies [
The Leibniz Science Campus “Digital Transformation of Research” (DiTraRe)1 [
ceur-ws.org
One of these use cases is Sensitive Data in Sports Science, within which the Motor Research data
repository (MO|RE)2 is being developed by the Karlsruhe Institute of Technology Institute of Sports and
Sports Science (KIT IfSS). MO|RE ofers possibilities for storing, publishing, and archiving data about
physical and cognitive capabilities generated during human motor performance testing [
Sports science data overlap with healthcare data. Existing approaches in related domains have so
far not provided suitable solutions to the challenges in research data for sports science identified in
DiTraRe. As such, ontologies like the Physical Activity Ontology (PACO) [
The MO|RE use case functions as a proof-of-concept for the approaches aspired in DiTraRe. This
concerns the modeling of complex research processes (e.g., standardized test items, measurement
procedures), as well as the design rationale suitable for encoding privacy aspects. This paper describes
the ongoing work done by the DiTraRe dimension “Exploration and Knowledge Organisation” within the
use case Sensitive Data in Sports Science. Our contribution focuses on developing a Basic Formal Ontology
(BFO)-based domain ontology and KG [
The management of research data has undergone significant evolution with the digital transformation of science. As already discovered around 20 years ago and shown in publications such as the work by Ludäscher et al. (2006), traditional approaches to the management of scientific data have proven insuficient for the complex heterogeneous datasets characteristic of modern-day research.
Semantic technologies have enhanced the research data lifecycle by providing formally grounded,
semantically explicit representations that capture domain semantics while facilitating cross-disciplinary
linking [
Upper-level ontologies, particularly the BFO [
With these features, the BFO provides coherent and logical representations, as well as explicit
formalizations. Ontologies that extend or build on the BFO ensure a high degree of interoperability with
other knowledge systems, enabling cross-disciplinary applications [
In a broader infrastructural context, the National Research Data Infrastructure (NFDI) initiative in
Germany brings together a diverse range of disciplinary consortia to create resources that improve the
management and sharing of research data across the sciences [
Further examples of applications, such as the PMD Core Ontology (PMDco), demonstrate the wide
range of uses for BFO. PMDco addresses key data management issues in materials science and
engineering. As a mid-level ontology aligned with BFO 2020, PMDco models core concepts related to material
classes, processing techniques, structures, properties, and performance characteristics, all of which
reuse existing patterns [
Similar to PMDco and NFDIcore, with the ontology for MO|RE data, we aim to harmonize complex measurement processes and contextual data for related test items.
Formal semantic approaches to organizing and representing sports science data remain underexplored.
However, ontologies in related domains provide formal representations of human motion and
physiological parameters. The OBO Foundry ontologies pertain to anatomy and physiology, enabling the
integration of biomedical data. For example, the Ontology of Biomedical Investigations (OBI) describes
processes that are also relevant for motor research [
The IfSS developed the MO|RE data repository as a platform for publishing and archiving empirical research data in the field of sports science. In the portal, users can upload and/or download anonymized, multi-dimensional datasets with (to a certain degree) standardized motor performance testing protocols, including the metadata. Data are published open access and with a DOI.
Derived from both long-term studies and shorter testing scenarios, the protocols capture diferent levels of human physical and cognitive capabilities across diverse experimental scenarios. In MO|RE data, a test item is a formalized procedure (e.g. Shuttle Run Test, Sit & Reach, 20 meter Dash), and is also described as a comprehensive assessment protocol. This enables the enactment of a concrete, measurable task with defined execution parameters, measurement criteria, and evaluation standards, thereby describing the necessary aspects for reproducibility.
Current data organization within MO|RE data follows traditional relational database principles. The datasets are provided as spreadsheet tables for the studies, together with test protocols. These tables include anthropometric data on participants (e.g., age, height, weight, and BMI) and measurement results for test items. However, the data in the form of relational databases makes it dificult to formalize the representation of complex temporal relationships and the interrelation between predefined test item structures, their executions, and the results from real-life tests.
As already mentioned, research studies in empirical sports science often involve complex relationships between predefined test items and action specifications, participants and their characteristics, as well as measurement procedures (how is the weight and height measured, and what are the contextual factors?).3 Essentially, each study available in MO|RE data represents a temporal process with multiple phases, which are condensed into a spreadsheet form as a result set. In addition to these results, MO|RE data provides metadata for each study and definitions for the test items in separate files.
Central modeling challenges include capturing the hierarchical structure of test items and their inclusion in a study. For example, when categorized accordingly in MO|RE data, a fitness evaluation can consist of multiple subcategories (strength, endurance, coordination, flexibility) for a performance profile of the participants. These categories are then divided into test items that aim to assess each of these qualities. Furthermore, these evaluations encompass a temporal dimension. This dimension may involve recording a specific time point within testing sessions in the dataset.
Nonetheless, another level of importance can be attributed to the anonymization and privacy of the
included data. Sensitive data in sports science is prevalent, especially when it comes to federations and
interconnection to other databases, such as KonsortSWD4. The latter provides information on social,
behavioral, educational, and economic status. Within the KG, both protected and non-protected data
should be represented equally, which necessitates implementing diferent levels of access to maintain
data privacy [
The BFO functions as upper-level ontology for MO|RE data. Essentially, it diferentiates between
continuants (entities that persist through time) and occurrents (processes that unfold over time),
which addresses the temporal complexity inherent in motor performance research, as represented in
the datasets from the MO|RE repository. Moreover, BFO defines qualities as entities that inhere in
their bearers, enabling the formal representation of measurable features of participants, such as their
height, weight, or results in fitness evaluations [
The Handgrip is a specific type of test item that participants undergo during evaluations to measure the maximum force that the hands can exert. As part of the measurement process, a hand dynamometer is pressed as hard as possible with one hand at a time. The resulting maximum force exerted by the hand is measured by the dynamometer and recorded as the measurement result.5
The vision for the first version of our BFO-based MO|RE ontology 6 focuses on the formal representation of studies, participants, and test items, aiming to capture the essential components of motor performance research within the repository.
Data featured in sports science research, such as the studies and associated test results MO|RE provides,
often cover a range of personal information, be it anthropometric values or longitudinal health indicators.
While the repository already anonymizes data communicated to the general public by reducing feature
3https://motor-research-data.de/en/Testitems_en.pdf
4https://www.konsortswd.de/
5https://www.ifss.kit.edu/more/english/248.php
6GitHub repository of the MO|RE ontology, https://github.com/ISE-FIZKarlsruhe/more-ontology
information, additional mechanisms are required in the ontology and KG to ensure compliance with
privacy regulations (e.g., the General Data Protection Regulation given by the European Union) [
As Figure 1 shows, research studies are modeled as instances of more:study, a subclass of iao:plan specification, which is a subclass of iao:information content entity. This design choice reflects the understanding that a research study is a structured plan, thus an information artifact specifying a sequence of actions or processes intended to achieve certain research objectives.
The same principle applies to specifying test items in the ontology. A more:test item is, same as the study, an iao:plan specification. These test items, in essence, are action specifications (e.g., exercises like “pushups” or “shuttle run”, represented as individuals) that define what participants are expected to do within the study. Each test item from a dataset is concretized in an iao:plan, which is accordingly realized in a concrete more:test process, a subclass of obi:assay and bfo:process. This is simplified with pato:executes.
Accordingly, the Handgrip test item is modeled as an individual of type more:test item. The results are captured as the output of a measurement process, modeled as an instance of more:handgrip test process, which realizes the specified plan. The actual Handgrip test process of a participant is modeled as an instance of a more:handgrip text process. Connected to it is the individual participant as an instance of more:person with the obi:evaluant role.
Study participants are modeled as instances of more:person, a subclass of bfo:material entity to represent their existence as continuant entities undergoing concrete assessments. All participants exhibit measurable qualities, such as anthropomorphic values measured before the evaluation (bfo:quality). Following the logic of the plan specifications, participants participate in the measurement processes with corresponding roles, particularly the obi:evaluant role. Roles are connected to processes via obi:has role and realized in relations. The dispositions inherent in persons are realized through the test processes. Their values are specified via the obi:has specified output property, which connects a process to the measurement datum (instances of iao:scalar measurement datum and corresponding subclasses). This measurement datum has a value specification ( obi:value specification), which, for the Handgrip example, is a decimal value in kilogram. This value is then used as a specification of the aforementioned disposition (obi:specifies value of). As Figure 2 shows, with a shortcut based on the previously defined relations between the resources, the ontology can also express the direct connection between a test item (the Handgrip in this case) and the disposition that it measures through the process (more:measures disposition) via a shortcut.
The same approach can be extended to other test items from the MO|RE repository, for example, a shuttle run, in which the more:test item would be a ex:shuttle run test item and accordingly, all connected entites would be adapted (more:shuttle run test process with instantiations like ex:shuttle run test process 1 for a specific execution of the test process). In the same sense, the dispositions are modeled in alignment with how a shuttle run measures the performance of a participant, e.g., measuring the time interval shown for a specific distance, or the VO2max value (maximum amount of oxygen a body can use during intense exercise).
This modeling approach enables a distinction between the planned Handgrip test, its concrete realization in a participant’s test process, and the result value(s). In a concrete application scenario, in which a sports scientist is interested in finding out how motor performance in elementary pupils has changed throughout the years between 2015 and 2020, the MO|RE ontology makes it possible to compare data from multiple motor performance studies, including Handgrip strength tests conducted across diferent age groups and over several years, which can then be compared despite variations in protocols. This can also be exemplified via competency questions (CQs), as applied for ontology development and evaluation in MO|RE. In this application scenario, such a CQ could be CQ1 ‘How does handgrip strength vary across age groups?’, or CQ2 ‘Which test items were included in studies conducted between 2015–2020?’ to filter out relevant triples. Both can be expressed as SPARQL queries over the MO|RE KG.
Listing 1: CQ1 expressed as a schematic SPARQL query for the MO|RE KG.
PREFIX more : < h t t p s : / / w3id . o r g / more # > PREFIX o b i : < h t t p : / / p u r l . o b o l i b r a r y . o r g / obo / OBI_ > PREFIX i a o : < h t t p : / / p u r l . o b o l i b r a r y . o r g / obo / IAO_ > PREFIX x s d : < h t t p : / / www. w3 . o r g / 2 0 0 1 / XMLSchema # > SELECT ? a g e (AVG ( ? s t r e n g t h V a l u e ) AS ? a v g S t r e n g t h ) WHERE { ? t e s t a more : H a n d g r i p T e s t P r o c e s s ; o b i : h a s _ s p e c i f i e d _ o u t p u t ? datum ; o b i : h a s _ p a r t i c i p a n t ? p e r s o n . ? datum a i a o : S c a l a r M e a s u r e m e n t D a t u m ;
o b i : h a s _ v a l u e _ s p e c i f i c a t i o n ? s t r e n g t h V a l u e .
? p e r s o n more : hasAge ? a g e . } GROUP BY ? a g e
ORDER BY ? a g e
The development of the MO|RE ontology and a corresponding KG is essential for the further development of motor performance test data within sports science, but especially for cooperation and collaboration with numerous external disciplines (health, education, psychology, etc.). Motor performance test data contain a wealth of information that has not yet been fully explored. The combination with other data from neighboring disciplines holds possibilities that are hardly foreseeable, for example, the relationship between motor competence and academic achievement, or the long-term health trajectories of children with low baseline motor skills. These questions require data infrastructures that go beyond singledomain perspectives. Importantly, we also emphasized the role of ontologies and knowledge graphs in promoting transparency, reproducibility, and reusability of data-core principles of open science. By structuring knowledge explicitly and making connections machine-readable, sports science stands to benefit from tools already well-established in bioinformatics and clinical research.
Additionally, based on the developments within the ontology, it is possible to define shortcuts and modules that can find application outside of this use case. This concerns shortcuts based on BFO models and, according to SWRL rules, those defined relations in particular. This plays a role in endeavors outside of DiTraRe, especially in related approaches, such as within the German National Research Data Infrastructure (NFDI), particularly in modeling complex processes and depicting data provenance.
In this paper, we introduce the vision of the MO|RE ontology, an ontology specifically designed for the needs of sports science and the MO|RE data repository. MO|RE is a platform for collecting, publishing, and sharing motor performance data. Our ontology, which is under development, is based on the BFO and reuses existing ontologies to build individual modules. The development of shared ontological frameworks in sport science is not merely a technical task; it is a strategic investment in the future of the ifeld. It enables interdisciplinary collaboration, supports evidence-based policy, and helps build bridges between scientific discovery and real-world application, particularly in areas such as youth development, public health, and inclusive education. The contribution presented in this paper is thus twofold: (i) to ontology engineering, it contributes a BFO-based module addressing processual modeling with privacy-aware extensions; (ii) to sports science, it provides a semantic infrastructure that standardizes motor test data.
The Leibniz Science Campus “Digital Transformation of Research” (DiTraRe) is funded by the Leibniz Association (W74/2022).
During the preparation of this work, the authors used DeepL and Grammarly for grammar and spelling checks. The authors reviewed and edited the content as needed and take full responsibility for the
As Figure 2 shows, with a shortcut based on the BFO-based relations between the resources, the ontology can also express the direct connection between a test item (the Handgrip in this case) and the disposition that it measures through the process (more:measures disposition) via a shortcut.