Principle-based category development for the digital humanities #

Overview #

The overarching goal of KatKit is to reduce the complexity of ontology design and support humanistic thinking through mathematically informed methods. This will reduce the risk of inadvertently creating incoherent and inconsistent models of the relevant objects of inquiry, while strengthening the conceptualization and operationalization of humanistic research questions.

The project aims to support the development of category systems in the digital humanities through mathematical principles. As a result, the ontology editor KatKit (KATegories toolKIT) will be developed as a proof of concept that 1.) is based on mathematical models that ensure properties such as consistency and coherence of category systems, 2.) supports humanistic reasoning, and finally 3.) can be integrated into existing DH toolchains, e.g. by connecting to the annotation software CATMA. In this project, the problem of category development is addressed in an interdisciplinary way by the digital humanities, philosophy, and the formal sciences of mathematics and computer science using applied category theory.

Problem definition #

Mathematical and computer science methods are constantly used in the digital humanities to process and analyze data. However, especially in the development of ontologies, tag sets, category systems, and classifications, which are a central part of scholarly work in the Digital Humanities, the use of formal scientific methods is not very pronounced. So far, genuinely digital humanities tools that support classification are not yet available. Category systems are often created in an unsystematic, ad hoc manner or are based on solutions from the natural sciences, life sciences, and engineering sciences, which can only be adapted to the requirements and procedures of the digital humanities with great effort.

Solution approach #

In order to remedy this deficiency, mathematically-informatically guided principles based on applied mathematical category theory will be developed to meet the needs of the digital humanities and their objects of research. The development of these principles and best practices forms the basis for an ontology editor that, on the one hand, supports humanities thinking and leads to appropriate modeling of the specific objects of investigation, and, on the other hand, produces robust results, i.e., ensures properties desirable for category systems, such as coherence and consistency.

Goals #

KatKit will open up the possibility of implementing a number of analyses in the digital humanities that have not been systematically possible until now:

• Supporting category-guided analyses. Annotatability of relations between individual instances of categories or integration of different levels of abstraction and thus avoidance of category errors.
• Development of category systems themselves. Modeling of cluster concepts and family similarities, multidimensionality or vagueness.
• Easy manipulation of category systems and their revision in the work process.
• Counterfactual play out of what-if scenarios.
• Modeling perspectives and contradictions in analyzed texts.

Funding #

This project is funded by the state of Hesse in the funding line “LOEWE-Exploration” from 12/2021 to 11/2023.

Contact #

Dominik Gerstorfer
TU Darmstadt
Residenzschloss 1 64283 Darmstadt