The Occupation Ontology (OccO) is a community-based ontology for occupations, which extends from the upper-level Basic Formal Ontology (BFO) in accordance with Open Biological and Biomedical (OBO) Foundry principles. In this article, we report on updates to core OccO definitions after expanding representational coverage to include codes from US SOC/O*NET and ESCO, as well as information from Wikidata. We also report on OccO working group collaborations with the ESCO team in the interest of integrating ESCO terms into OccO, and the Alabama ontology project team in the interest of representing competencies and credentials. We close by highlighting our strategy for integrating Wikidata and occupation standards using OccO as a lingua franca. We outline various updates and challenges throughout and encourage further participation from the community as we pursue this effort.
The Occupation Ontology (OccO) [
We intend to leverage OccO as a common framework for distinct coding standards. US SOC, UK
SOC, ISCO, and ESCO standards exhibit significant overlap in coverage, though they differ with respect
to representations of occupation: metadata attributes, skills and abilities, as well as descriptions of
associated tasks. Traditionally, mappings – or crosswalks – between occupation standards have been
created to promote interoperability. Crosswalks become, however, outdated following occupation
standard updates, often require significant manual curation, and susceptible to the “order n-squared”
problem [
Recently, NLP strategies have been deployed to address some of these limitations. Arguably, it
should not matter how many crosswalks are needed if they can be automatically generated and
maintained. For example, a 2022 EU Commission technical report [
These observations have demonstrated that there is a clear, pressing, need for constructing a comprehensive occupation ontology that harmonizes existing standard language taxonomies, represents major occupations, and features such as skills, abilities, credentials, and competence, and that facilitates bridging across occupation coding standards. Accordingly, OccO is designed as a semantically rich lingua franca into which occupation standards and occupation data can be mapped to promote interoperability and avoid the n-squared problem. To that end, we report on ontological refinements of key OccO terminological content, as well as our successes and outstanding challenges associated with incorporating other occupation terminologies within its purview. More specifically, we detail subtleties that emerged in key OccO definitions when placed in conversation with additional occupation standards, provide an updated design pattern for core OccO classes and relations, expand on our previous treatment of skills and abilities, and report our use of Wikidata to identify potential coverage issues for ESCO and US SOC and enrich OccO. Lastly, we outline challenges encountered, and invite researchers interested in collaboration to aid in efforts to build a bridge across international occupation standards.
OccO is designed as an extension of the Basic Formal Ontology (BFO), which is a top-level ontology
covering general classes such as material entity, quality, process, function, and role [
Instances of disposition are realizable entities - entities that if realized, are realized in processes; they are said to be “internally grounded”, a metaphor meant to track the fact that were the instance to cease to exist, then its bearer would be physically changed [18]. For example, if a piece of sodium chloride is soluble at some time, then not soluble at another, there must be some change to its physical makeup. In this sense, instances of disposition are not optional for bearers. Realizations of instances of disposition occur, moreover, owing to the material constitution of the bearer – and its associated qualities - and the fact that the bearer is in some environment the material bearer is not always in. The realization of salt’s solubility occurs owing to the lattice structure and bonding forces of the salt when placed in unsaturated water. The class role is a disjoint sibling class of disposition, and so a realizable entity, but one with characteristics sharply distinguishing it from its sibling. In contrast to disposition, instances of role are optional in the sense that bearers may gain or lose them without physical change. A student, for example, who graduates from a university no longer bears the role of student at that institution. That does not, however, imply any physical change in the student. More generally, whether an entity bears a role depends largely on what happens external to the entity.
Natural language and relevant occupation standards suggest “occupation” is often used in either a disposition or role sense. On the one hand, “occupation” may be used synonymously with “job”. For example, when asked to describe one’s occupation, it is not uncommon to hear responses such as “cashier” or “physician” or “professor”. As another example, the ISCO defines an occupation as "[T]he kind of work performed in a job" and 'job' as "[A] set of tasks and duties performed, or meant to be performed, by one person, including for an employer or in self-employment" [19]. A job is most naturally understood as a type of role in BFO, as individuals may gain or lose jobs without necessarily undergoing material change.
On the other hand, “occupation” may be used to describe a set of capabilities one has independent
of whether they also have a job. For example, a Python developer in between jobs may be described as
having a developer occupation, despite not being employed. Indeed, Classification Principle 2 of US
SOC states that “Occupations are classified based…in some cases, on the skills, education, and/or
training needed…” to perform a job [
We maintain that the use of “occupation” should contrast with the processes associated with occupations. An employed Python developer is authorized to participate in code review for an organization; an unemployed Python developer has a skillset manifested in repository updates, opensource contributions, etc. This position contrasts with, for example, the American Occupational Therapy Association which defines occupation as "Various kinds of life activities in which individuals, groups, or populations engage, including activities of daily living, instrumental activities of daily living, rest and sleep, education, work, play, leisure, and social participation" [21]. Besides such a definition being so broad as to include nearly all human activity, applying this definition to occupation standard data would conflate jobs, occupations, and activities associated with either.
There are identifiable commonalities across the above uses of “occupation”. Occupations are inherently transactional. Employers seek talent to fill job and unemployed individuals having occupations often seek employment. This should not suggest that anyone hired for an occupation is thereby compensated for work performed. Employees should be compensated, but employers do not always behave as they should; an unpaid employee who deserves compensation may count as holding an occupation in the job title sense, nevertheless. Similarly, individuals holding occupations may not perform corresponding duties or perhaps may perform them poorly. We do not say, however, that a poor performer does not hold the relevant occupation.
This leads to points of ontological complexity worth belaboring: competency levels and relevant
intentions. Consider Chris, a cashier at a local grocery store who – despite lacking training or having
previously expressed interest in doing so – decides he has a computer programmer occupation and
intends to develop the skills needed to obtain a programming job. It is implausible to count Chris as
having a computer programmer occupation simply because of his stated declaration; he needs to do
more to be counted as having such an occupation. Though we perhaps cannot draw a firm line between
when one has sufficient relevant skill to count as having an occupation and when one does not, we can
point to paradigmatic cases where individuals clearly have an occupation and where not. In this
example, Chris lacks relevant skills. That said, Chris’s intention to develop relevant skills and seek
employment, seems an important component to having an occupation. Despite being a skilled cashier,
after Chris quits his cashier job and decides to no longer pursue cashier jobs, he should not be counted
as having a cashier occupation. Worth noting, the ISCO definition of “occupation” supports the
importance of intentions to having an occupation, stating that occupations are associated with activities
performed and activities intended to be performed [
The preceding hallmarks were used to refine OccO definitions and construct additional terminological content needed to pursue our goals of bridging occupation standards.
Figure 1 illustrates the core OccO design pattern, which has been re-engineered to reflect the preceding hallmarks of occupation. We import the class human from the NCBITaxon [22] as a bridge between the BFO class material entity and those who hold occupations. An occupation holder is someone bearing either an occupation role or an occupation disposition. This class captures the bearers of occupations, such as a pharmacist, welder, ontologist, etc. The substance of this description of occupation holder is found in the definitions of occupation role, occupation disposition, skill, and ability.
The class occupation role is meant to capture the “job” sense of occupation. Bearers of such roles provide labor or services in exchange for compensation, and the existence of such a role requires an authority capable of empowering the bearer to perform relevant tasks for compensation. Individuals are often hired to perform tasks based on perceived ability and skill to do so. Accordingly, OccO introduces ability and skill, which fall under disposition. Regarding the former, ability represents dispositions - naturally acquired during an individual’s development – to perform tasks, e.g., walking, jumping, eating. As such, ability is a rather broad class. Even so, abilities relevant to an occupation are bound by the scope of duties associated with that occupation. A surgeon may have the ability to jump, but a hospital employer would likely not find that as interesting as, say, being able to communicate with patients. Regarding the latter, skill represents dispositions emerging from training that involve a high-level of proficiency and which can be exercised reliably. The hospital employer would likely find a potential surgeon’s dispositions to successfully complete open-heart surgery, placing a stent, etc. to be of significant interest. Training, reliability, and proficiency distinguish skill from ability, though both are needed to characterize occupation disposition.
We treat skill and ability as siblings, which may seem odd since intuitively skills are simply abilities that have been honed through training. However, classifying skill as a subclass of ability would entail that any instance of skill counted as an instance of ability, and it seems mistaken to say that in every case once one cultivates a skill, they thereby gain an ability. First, developing a skill involves many abilities; there is a many-one relationship between abilities and skills. Second, developing a skill should not entail that one thereby gains a disposition of the sort naturally acquired during an individual’s development. Third, treating skill as a subclass of ability runs the risk of suggesting that for every skill there is some ability that was trained into that skill, which is false. One does not become an expert programmer by training up a native programming ability. Consequently, we do not classify skill as a subclass of ability.
The class occupation disposition is defined partly in terms of some ability relevant to an occupation role and partly some skill aimed at performing tasks associated with that role reliably and proficiently. However, skill and ability are insufficient alone to capture the hallmarks of this sense of occupation. One may have skills and abilities relevant to a job but have no intention to work in such a position. Thus, to adequately characterize occupation disposition we must appeal to the intentions of bearers to pursue work of a sort relevant to an occupation [23]. Discussion of the precise nature of intentions is notably complex and beyond the scope of our work [24]. For our purposes, in general we understand intention as the directedness - or aboutness towards - of some cognitive process [25], and we understand intention as the directedness of a bearer’s mental representation towards pursuing work relevant to an occupation. In short, occupation disposition is defined in terms of relevant abilities, skills, and intentions of a bearer. Lastly, occupation holders participate in instances of occupation activity, which are - roughly - processes in which such participants engage in pursuits associated with associated job titles or intended career paths. Definitions of key OccO terms are found in Table 1.
A human bearing an occupation role or occupation disposition.
occupation role occupation disposition ability
A role borne by a human that, if realized, is realized when the bearer provides labor or services in exchange for compensation as specified by some deontic declaration.
A disposition that, if realized, is realized when the bearer intends to, and does, exercise abilities and skills in pursuit of obtaining or maintaining an occupation role.
A disposition that inheres in a human, exists in virtue of natural biological development, and is such that if realized, is realized in the performance of one or more tasks.
skill occupation activity
A disposition that inheres in a human, exists in virtue of training, and is such that if realized, is realized in the performance of one or more tasks with reliable high proficiency.
A process in which an occupation holder participates that realizes either an occupation disposition or occupation role.
The Alpha version of OccO mapped the US SOC/O*NET occupation categories and detailed occupation labels, but our aim is to integrate other occupation categories under the umbrella of OccO. For example, the ESCO taxonomy overlaps significantly in terminology with US SOC, thereby facilitating mapping to OccO. Even so, integrating ESCO with OccO representations of US SOC/O*NET is challenging. ESCO shares the same 10 Major Groups with ISCO, represented with a single digit code. In contrast, US SOC includes 23 Major Groups. While both ESCO and US SOC/O*NET use a four-layer hierarchy of intermediate classes, in each case “dotted suffix” codes are used to extend the ISCO and US SOC/O*NET codes. An example is “barista” where the ESCO code is 5132.1.1 using the four-digit ISCO code and the corresponding O*NET code is 35-3023.01, using the SOC seven-position code. It is important to preserve the existing group categories because many reports are summarized by major or minor group.
As discussed in the introduction, there have been recent attempts to employ NLP strategies to automate crosswalks between occupation standards. One result of this growing interest has been “Dynamic Ontology Matching Challenge” [26], a community call to action to automate – to the extent possible, while acknowledging there will likely need to be significant manual labor involving in such strategies. The OccO development team has had several meetings with ESCO team members Gianluca Bortoletto and Jan Luts – authors of the NLP technical report outlined in the introduction - over how OccO might be employed to address this challenge. To that end, we propose a schema for connecting codes from distinct occupation standards along the following lines, using “barista” as an example: OccO:barista OccO:has_ESCO_code “5132.1.1”^^xsd:string ;
OccO:has_O*NET_code “35-3023.01”^^xsd:string .
In words, for a given category common between these occupation standards, we introduce a relevant occupation code datatype property with the corresponding literal value. In this way, the respective codes are linked in OccO by – in this case – the OccO class barista.
Exact matches are straightforward, but do not address all mapping challenges. In many cases, a term in ESCO is a broad or narrow match to an O*NET term; in other cases, the relationship between occupations in these respective standards is unclear. Our current strategy for capturing exact matches in OccO is illustrated above with barista. For close and inexact matches, we employ the following strategy:
OccO:Example1 OccO:has_ESCO_code "###"^^xsd:string ;
OccO:has_O*NET_inexact_match OccO:Example2 .
OccO:Example2 OccO:has_O*NET_code “"$$$"^^xsd:string ;
OccO:has_ESCO_inexact_match OccO:Example1 .
In other words, the term XYZ which bears the ESCO code ### also bears an inexact match relationship to OccO:IRI123, which bears the O*NET code $$$. Additionally, because ESCO and O*NET provide descriptions of relevant abilities and skills for these occupations, OccO can facilitate semantic representation of the similarities and differences between them. Worth stressing is that the preceding proposal is tentative, as we are considering adopting the SSSOM mapping standard [27]. 4. OccO Integra=on with the Alabama Talent Triad A program related to OccO is being conducted by the Alabama Committee on Credentialing and Career Pathways (ACCCP), with the leadership from the Governor’s Office of Education and Workforce Transformation (GOEWT). This group has developed an informal Alabama occupation ontology as part of its Talent Triad initiative, which aims to connect individuals, employers, and credential providers using a common skills-based framework. As its name suggests, the Talent Triad has three components: Skills-Based Job Descriptions, Learning and Employment Records, and Skills Defined Credentials [28].
The Talent Triad pilot phase [29] will be accessible through a software application that relies on socalled “Skills DNA” to support algorithmic matches between individuals seeking employment and employers. The Skills DNA model relies on an informal ontology of skills and/or competencies [30]. Alabama’s ontology took as its starting point the Building Blocks Model developed by the Competency Model Clearinghouse within the Employment and Training Administration of the U.S. Department of Labor, which organizes competencies in a tiered structure with increasing specificity, culminating with occupation-specific competencies. Through work with experts including Corporation for a Skilled Workforce [31], Alabama’s ontology has made several enhancements, such as grouping skills/competencies into standard “functions” that can be contextualized to specific occupations. Approximately 350 high-demand occupations from US SOC/O*NET are targeted.
In addition to matching individuals with job opportunities based on skills they currently possess, the Talent Triad will suggest additional education and training that an individual can pursue to increase skill level or acquire new skills. This feature relies on credential information being organized in a standardized way around skills and/or competencies. Alabama, along with over 20 other states, has partnered with non-profit organization Credential Engine [32, 33] to represent credential information using the Credential Transparency Description Language (CTDL) [31], written in OWL. CTDL allows for the representation of key descriptors for credentials, including the organization offering the credential, its requirements, and alignment with specific occupations. In addition to serving as the clearinghouse for the development of CTDL, Credential Engine maintains an extensive database of hundreds of thousands of credentials known as the Credential Registry, and Alabama will continue to publish credential information from the Talent Triad to the registry maintained by Credential Engine.
A key element of the Alabama program is the articulation and development of skills and abilities associated with the target occupations, described as credentials and competencies. While under development, the OccO working group is considering the following representations of such phenomena: • Following [34], credential can be understood as a document that reflects an attestation to the holder’s competence with respect to a given task, at or above a certain threshold. Accordingly, credentials are issued by some authority and are numerous. • Related, competence can be understood in terms of measurements of the degree to which the combination of skill and/or ability has been developed. Accordingly, competence deals with evaluations of an individual’s ability to perform a job.
By focusing on skills, abilities, credentials, and competence, developers of ACCCP aim to unite employers, educators, and members of the workforce with a common ontological language. In this respect, the aims of ACCCP align with the aims of OccO, as we too seek to integrate and enrich occupation data using standardized language. Future development of OccO will seek to align not only our aims, but also our respective understandings of these phenomena.
As a concrete illustration of how we understand OccO’s utility, we turn to Wikimedia resources, such as Wikipedia, Wikidata, and DBpedia. Wikipedia contains significant quantities of occupation data as well as references to occupation holders. DBpedia is a linked open data project aimed at extracting structured content from Wikipedia to be placed in a triple store for semantic querying and analysis. Wikidata is a vast semantic web resource used to generate linked open metadata as a supplement to Wikipedia pages. Wikipedia is used primarily as a crowd-sourced publicly editable information storehouse. DBpedia, on the other hand, is used more often by researchers, and is updated infrequently. Wikidata may also be edited by users directly. Consequently, Wikidata needs curation, hence our focus on this Wikimedia product for our OccO use case.
For example, we retrieved a total of 4,617 people from Wikidata who died from COVID-19 as of January 2022. One might be interested in investigating the classes of occupations represented among this population, perhaps to explore occupation risks associated with the disease. Such analyses would, however, require robust semantic structure, which Wikidata does not exhibit. Of the 4,617 with both date of death and one or more occupations, the result is 4,007. Figure 3 shows the number of deaths per month for people in Wikidata through January 2022.
Data about individuals in Wikidata exhibits a triple structure, where resources are connected via properties. Two important identifiers and labels are listed in Table 2.
Identifier
Label occupation has occupation # of Occurrences ~3000 ~17000 Of the occupation relation occurrences in Wikidata, only a quarter of these match a resource classed as a Q12737077 occupation or descendant. Moreover, perusal of occupations in Wikidata revealed that only 772 of its occupation labels are associated with ISCO-2008 occupation classes, while 397 are associated with ESCO occupation identifiers, 600 with ESCO skill identifiers, 112 with US SOC identifiers, and 85 with O*NET codes. Each Wikidata label is associated with a specific number of occupation labels in the Wikidata taxonomy, as illustrated in Table 3. Notably, the listing of ISCO occupations in Wikidata is complete, following a comprehensive update of ISCO resources, such as breaking up terms such as “mathematician and actuaries” into individual occupations.
Wikidata Label
Wikidata Code ID # of Labels Matching Occupation Code ISCO-08 occupation class ESCO (V1) Occupation ID
ESCO (V1) Skill ID
SOC ID O*NET
P8283 P4652 P4644 P919 P8734 772 397 600 112 85
Wikidata lacks ontological rigor despite exhibiting triple structures. This makes integration with
Wikidata challenging, if pursued in a systematic fashion. For example, in Wikidata, the class
“occupation entity” (Q12737077) is an instance of two parents: "concept" (from which all its entities
descend) and "second-order class" (a term that we have not yet found a formal ontology using). The
entity "profession" (Q28640) is an instance of "Wikidata meta class" and "second-order class".
Therefore, querying for all instances of “occupation” will miss the many occupations that are instances
of “profession.” More specific examples are also provided in our previous OSS2022 paper [
Connecting Wikidata to other occupation standards will be made easier by using OccO. Our strategy – illustrated in Figure 4 – is to represent distinct occupation terminologies within the umbrella of OccO. It is our belief that such an approach will lead to the achievement of data integration across occupation standards. Roughly speaking, the strategy is to use OccO as a lingua franca. This allows for the enrichment of these standards with ontological structure; such enrichment facilitates the use of a query language like SPARQL to run quality control over datasets ontology, avoid scope creep, all while maintaining a common language into which each standard can be translated to promote interoperability.
A major goal going forward will be to implement the above strategy while attempting to apply some formal structure to domains. By doing so, we will then be empowered to analyze individuals with different occupations or occupation groups, and their associations with different scenarios such as death due to infectious diseases like COVID-19.
We have described several contributions in this article. First, we formed an OccO working group consisting of researchers from a wide variety of academic institutions. Second, we achieved consensus on the definitions of occupation-related terms within the BFO/OBO framework. Third, we introduced and developed our framework in discussion with representatives of similar efforts in the ESCO and Alabama GOEWT, in the interest of benefitting from subject-matter expertise and promoting stakeholder investment. Fourth, we investigated the use of the OccO ontology in studying Wikidata resources of various disintegrated occupation types and individuals.
Next steps include incorporating ISCO and ESCO occupation titles into OccO, after which work may begin on harmonizing ESCO skills with corresponding O*NET skills and abilities, as each set of skills will be represented in OccO and associated with occupation classes. Additionally, our work will focus on achieving consensus on ontology content needed to represent specific skills, abilities, credentials, and competencies. Given how important skills and abilities are to US SOC/O*NET and ESCO, and how important credentials and competencies are to many relevant stakeholders working in this space, incorporating this content into OccO is of crucial importance.
Our collaboration between the OccO project and Alabama project will bring many benefits and applications. OccO will bring these skill-ability-educational properties into some degree of alignment or uniformity. It is possible to combine O*NET skills and abilities properties along with Credential Engine properties to harmonize these important occupational properties. It would be a worthwhile exercise to see to what degree credentials that are now widely diverse could be organized into a more orderly and common structure, using CTDL and OccO.
The usage of the three Wikimedia resources allows us to access thousands of persons with specific occupation types, which may be standardized and integrated using our ontological strategy. However, our study found that the classification of occupations in Wikidata is uncurated and confusing. Therefore, we propose to use OccO for standardizing and integrating these resources, which will significantly improve our study of occupation related instance data and allow us to address various occupation related questions. One further issue worth noting that stems from working with Wikidata is that because this data source is ultimately crowd-sourced, it may contain false or misleading information concerning occupations. We envision that stakeholder engagement, subject-matter expertise, and careful representation of Wikidata occupation content in OccO, will provide a crucible in which to uncover inconsistencies, identify misinformation, and correct it through careful analysis. In this respect, the OccO workflow will act as a line of defense against such misinformation.
Many future directions are possible for our OccO working group research. The ISCO is updated every twenty years, with the next revision in 2028. We believe the next version would benefit from OccO development. We also expect that the OccO ontology effort will be a part of the Behavioral and Social Sciences Ontology (BSSO) Foundry, which has been formed to address the extensive array of social science-related information. The OccO workgroup is evaluating whether to join forces with this activity for our next stage development, in its current informal status, but the hope is that OccO activity can be elevated to a more formal, endorsed, and funded activity by one or more occupation standard organizations as this step was taken.
We would like to thank Gianluca Bortoletto, Jan Luts, and the ESCO team managed by the European Commission; our infrequent meetings were informative and invigorating. We would also like to thank our correspondents in Alabama, Nick Moore, head of the Governor’s Education and Workforce Transformation project. Additionally, we would like to thank Fumiaki Toyoshima and Adrien Barton for helpful feedback on an early draft of this paper. [18] Arp, R., Smith, B. "Function, Role, and Disposition in Basic Formal Ontology." Nat Prec, 2008, https://doi.org/10.1038/npre.2008.1941.1. [19] "Concepts and definitions." International Labour Organization, Accessed 6 June 2023, https://ilostat.ilo.org/resources/concepts-anddefinitions/classificationoccupation/#:~:text=and%20ISCO%2D88. [20] "Labor Force Statistics from the Current Population Survey." U.S. Bureau of Labor Statistics,
Accessed 6 June 2023, www.bls.gov/cps/cpsaat25.htm. [21] American Occupational Therapy Association. "Occupational Therapy Practice Framework: Domain and Process (3rd ed.)." American Journal of Occupational Therapy, 68(Suppl. 1), S1–S48, 2014, https://www.aota.org/~/media/Corporate/Files/Advocacy/Federal/coding/OT-PracticeFramework-Table-1-Occupations.pdf. [22] "Taxonomy Browser." National Center for Biotechnology Information, Accessed 6 June 2023, https://www.ncbi.nlm.nih.gov/taxonomy. [23] Hastings, J., et al. "Representing Mental Functioning: Ontologies for Mental Health and Disease."
Proceedings of the Third International Conference on Biomedical Ontology, 2012. [24] Toyoshima, F., Barton, A., Grenier, O. "Foundations for an Ontology of Belief, Desire and Intention." 11th International Conference on Formal Ontology in Information Systems (FOIS 2020), Sep 2021, Bolzano, Italy, pp.140-154. [25] Kasmier, D., Limbaugh, D., Smith, B. "Foundation for a Realist Ontology of Cognitive Processes." Proceedings of the International Conference on Biomedical Ontology, University at Buffalo, NY, 2019. [26] de Boer, M. H. T., et al. "Dynamic Ontology Matching Challenge." TNO - Netherlands Organisation for Applied Scientific Research; Universiteit Leiden, Accessed 6 June 2023, https://www.example.com/dynamic-ontology-matching-challenge. [27] Matentzoglu, N., et al. "A Simple Standard for Sharing Ontological Mappings (SSSOM)."
Database, vol. 2022, 2022, baac035, https://doi.org/10.1093/database/baac035. [28] "EBSCO Education Solutions and Services." EBSCO Information Services, 2023, Accessed 6 June 2023, https://solutions.ebscoed.com/. [29] "Building Blocks Model." CareerOneStop, U.S. Department of Labor, Employment and Training Administration, Accessed 6 June 2023, https://www.careeronestop.org/CompetencyModel/CompetencyModels/building-blocksmodel.aspx. [30] "Competencies & Credentials." National Network of Business and Industry Associations, Corporation for a Skilled Workforce, Accessed 6 June 2023, https://skilledwork.org/what-wedo/competenciescredentials/. [31] "Federal, State, and Regional." Credential Engine, Accessed 6 June 2023, https://credentialengine.org/credential-enginepartners/federal-state-and-regional/. [32] "Credential Transparency Description Language (CTDL) Handbook." Credential Engine, Accessed 6 June 2023, https://credreg.net/ctdl/handbook#credential. [33] "Alabama." Credential Engine, Accessed 6 June 2023, https://credentialengine.org/partners/alabama/. [34] Smith, B., Loddo, O. G., Lorini, G. "On Credentials." Journal of Social Ontology, 6(1):47-67, 2020.