Armin Gander from TU Wien will present an online talk within the seminar “Analysis of Complex Systems” on April 23, 2021, 3–4 pm (CET) via Zoom.
If you would like to attend, please email office@csh.ac.at
Abstract:
The phenomenon of colexification describes occurrences in natural language when two concepts are expressed by the same word in at least one language. We deploy this linguistic principle to construct a theory-driven text analysis method. Compared to many state-of-the-art Natural language processing (NLP) models, this method is fully interpretable, allowing precise insights into the structure of the model. Such theory-driven approaches are increasingly in demand since large NLP models make it difficult for developers to understand the models’ dynamics and implications thereof.
Using the text analysis method developed in this work, we validate the hypothesis that colexification occurrences encode semantic relationships between concepts and show that a colexification-based approach to NLP has significant merits in various text analysis tasks. For instance, our experiments show that the style of English fiction literature has become more diverse over time, with the rate of change increasing particularly sharply in recent decades.