Delivering collaborative solutions for standardised ontology development
Details
The collaborative development of a shared or standardised ontology presents unique issues in terms of workflow, version control, testing and quality control. These challenges are like those faced in large-scale collaborative software development, but no integrated platforms were available to handle them.
Our solution includes the following key components:
- customised GitHub repository integration. We created a specialised GitHub repository to manage the complexities of ontology development. The repository facilitates the smooth control of versions, allowing multiple contributors to work on different aspects of an ontology simultaneously. The system also tracks changes meticulously, ensuring that every update is logged and can be reviewed and rolled back if necessary;
- Jenkins continuous deployment/continuous integration (CD/CI) process. We integrated the Jenkins CD/CI pipeline with the platform, a process that automated testing, deployment and QA. As a result, every ontology change is rigorously evaluated before being merged into the main branch. This not only speeds up the development cycle, but also minimises the risk of errors, ensuring that an ontology remains consistent and reliable;
- a user-friendly web application for ontology end-users. Developers are not the only end users of ontologies, so we have developed a web application that allows easy access for a wider audience that includes non-technical users. It provides intuitive tools for searching, exploring and using an ontology and it also allows users to propose changes, which can then be reviewed and integrated by the development team;
This platform is now deployed for four ontological projects:
FIBO is a formal ontology that provides a common vocabulary for financial contracts and related concepts. It evolved out of concerns that arose during the 2008 financial crisis among people working together in data governance and management to address requirements for standardised terminology for regulatory reporting and other analyses.The result was an ontology that has continued to grow and evolve since its initial publication in 2014, with increasing support for use cases related to securities master data management, reporting and risk analysis and management.
The President of the Pistoia Alliance, Dr Becky Upton, describes IDMP-O as:
“(...) one of the Pistoia Alliance’s flagship projects with nearly a dozen of our large member companies involved. Our ontology bridges the gap between regulators, pharma, manufacturers, and the clinical space by offering a powerful common language for the description of medicinal product information. With IDMP-O, organizations can more easily comply with standards set by organizations like ISO, reduce duplicated efforts, and ensure patients are prescribed the correct medication for their needs. This project’s success has been made possible by the Alliance’s legal framework for pre-competitive collaboration, and underlines the value of the Alliance to the life sciences sector”.
Source: Pistoia Alliance website
Automotive Ontology (AUTO) is an OWL ontology based on the auto schema.org extension from the W3C Automotive Ontology community group. It defines key concepts in the automotive industry, highlights important vehicle-related objects to serve as a foundational resource for Knowledge Graphs and improves machine learning applications.
The IOF is part of a non-profit organisation, Open Applications Group (OAG), that aims to lower integration costs and meet the challenge of a swiftly evolving global digital economy by developing not only standards that are interoperable, cross-functional, cross-industry and driven by data-models, but also extensible standards.
The purpose of IOF is to create a suite of ontologies designed to support digital manufacturing by facilitating cross-system data integration in a number of spheres, namely, within a factory and across a business; in commerce between suppliers, manufacturers, clients, customers and other trading partners; and throughout a product’s life cycle. The IOF’s ontologies consist of ‘Core’, a common mid-level ontology, and several domain specific ontologies.
From an architectural perspective, the IOF Core Ontology, which contains terms present in a number of manufacturing operational spheres crowns this suite. The architectural approach selected by the IOF involves rooting all of its ontologies in one foundational or top-level ontology. Here, the Basic Formal Ontology, or BFO, was chosen. A great many of the intermediate-level terms in the Core Ontology derive from the BFO and the IOF ontologies draw on for them for a multitude of domain industry terms. Core intermediate-level terms are often domain independent. This means that they are present in other industries and fields like the banking, insurance and healthcare industries. They can also be found in the sciences, including the physics, chemistry, and biology domains.
Source: https://spec.industrialontologies.org/iof/