Expected learning outcomes
- Understand that change cannot be modelled on static vector space.
- Understand that as dynamics can be modelled conveniently by a vector field, semantic change needs a field model for its proper study.
- The ability to problematise about the implications, e.g. how to match dynamic semantics with evolving fields.
What is the target audience?
- Digital Preservation and Digital Humanities practitioners who are interested in investigating the deployment of semantics and relevant cutting-edge technologies.
- Researchers working on semantics who are interested in investigating the application of relevant methods and technologies in a Humanities field.
- Students in computer linguistics and knowledge engineering interested in evolving semantics.
Level of advancement/ prerequisites
This recorded lecture is part of a larger module “Dynamics of knowledge organisation”: http://pericles-project.eu/training-module/dynamics-of-knowledge-organisation/
This is chapter 3 of the complete set:
- Semantics and DP: basic concepts, theories and trends
- Vectors and matrices: word meaning for advanced access
- Vector fields: a new approach to evolving semantics
- The Semantic Web & the Emergence of Ontologies
- Semantic Technologies & DP
- Evolving Semantics: Ontology Evolution & Semantic Drifts
Time required for completion24:21 min
Wittek et al. discuss how the requirement to model semantic change leads to lexical fields as a preferred model of word meaning; how evolving lexical fields presuppose a vector field for representation instead of vector space; and how Emergent Self-Oganizing Maps as the mathematical approach meet Aristotle’s ideas about actuality and potentiality as two components of a continuous reality.
Wittek, P., Darányi, S., Liu, Y-H. 2014. A Vector Field Approach to Lexical Semantics. In: Proceedings of 8th International Conference on Quantum Interaction, Filzbach, Switzerland. June 30 – July 3, 2014.
Ultsch and Moerchen’s report introduces two ideas to model evolving vector fields: to model the tension structure vs. the content structure (i.e. content distribution) of a high-dimensional dataset by its mapping to a 2-dimensional representation.
Ultsch, A., & Moerchen, F. 2005. ESOM-Maps: tools for clustering, visualization, and classification with Emergent SOM. Technical Report Dept. of Mathematics and Computer Science, University of Marburg, Germany, No. 46.