The IM worldview: a shared knowledge base for science
This page contains a brief outline of its final contents, which will be completed in 2019.
- In theoretical work, we have suggested that the boundaries of a cross-domain interoperable conceptualization for integrated modelling are set by a choice of observational scale. Within these boundaries, a set of terms can be identified that can provide unambiguous and coherent building blocks for the dialogue between different disciplines. We name such a conceptualization a modelling worldview; semantic modelling tools can be used to build manageable worldviews that are more understandable to users than the current state-of-the-art formalisms, and integrate community-endorsed vocabularies and thesauri.
- Building ontologies is notoriously time-consuming and difficult, largely because different disciplines make different use of the same language. With the integrating principles of semantic modelling specified in the k.IM language, we are confident that this effort can be more productive than many past attempts. The IM partnership can thus play an important role in ensuring representativity, tightness, cross-domain endorsement and generality for semantic modeling.
- The IM partnership maintains the IM worldview as a set of k.IM-specified ontologies built on our general conceptualization of observables and universals, and geared toward the description of phenomena and problems related to Earth and its natural and human inhabitants. More general ontologies (describing basic aspects of the physical world) are specialized into core science domains (chemistry, earth, biology), which are the basis for more specific domain ontologies (agriculture, ecology, soil science, economics, policy, behavior). The most specialized ontologies in the worldview express narrower, but still mainstream, views of the sciences (such as ecosystem services), implemented as roles. Rather than introducing new concepts that would overlap with existing physical domains, roles describe viewpoints, reinterpreting those physical concepts in a particular light, without ambiguity or confusion. The validation infrastructure built into the toolset supporting k.IM ensures that all these different categories of knowledge are used in a proper and harmonious way, guiding the user towards consistent and minimal specifications.
- The IM worldview is shared across all the semantic web servers maintained by the Partnership and all partners. It is tied to the IM certificate and automatically synchronized so that each user is guaranteed to be always using the latest version.
- We commit to ongoing development and maintenance of the IM worldview, a long-term endeavor directed by all partners and users. This should not prevent other groups from using our work to develop differently scaled worldviews – for example for particle physics or cosmology. As each user and network server is committed to a worldview through the IM certificate, entire new networks can be set up with the k.LAB software stack without risking interference or confusion. Our aim for the IM worldview is to cover all the domains that are “resolved” from the natural observational scale of a human observer, encompassing natural and social sciences as well as their integration.
- Our effort gives a first possibility to provide a practical cross-domain conceptualization, small enough to learn and use by a diverse community of non-specialists, through rigorous modularization and subdivision of content. Worldviews strictly define the what of information (observable semantics), and as such do not need to address context (semantics of time and space), provenance or observation semantics. By separating universals from observables and allowing users to compose them freely in k.IM, we can greatly reduce the number of concepts. Using authorities, we can communally endorse and link external, large vocabularies of identities (e.g., taxonomic names or chemical species) to the k.IM specifications without having to incorporate those terms in the worldview. This further enhances usability and modularity when specifying semantics.
- Development of the worldview and its timeline. Quick description of the community process and workflow enabling use of provisional concepts. Concepts are resolved to the worldview through a collaborative process, when required semantics cannot be found – because concepts have not yet been defined or because established ideas about existing concepts are hard to reconceptualize.
- Use of the worldview for annotation, automatic documentation and discovery through crawlers.