This thesis revolves around two primary goals: First, an investigation into the linguistic properties that distinguish metaphorical usages from non-metaphorical ones and the formulation of quantitative descriptions of these properties. Second, the design and implementation of software to provide information to annotators and users more generally about the nature of semantic shifts that tokens undergo relative to the general usage of the words involved, especially with respect to potentially metaphorical usages.
Since this project deals with historical language variants and the phenomenon under investigation, metaphor, involves a challenging and complex annotation process, there will never be large amounts of data for us to work with. Furthermore, since the goals of this project are primarily scientific in nature, the methods employed in the project must be such that they allow us to generate new and reliable knowledge about our problem domain. Therefore, widespread and now-conventional approaches based on large neural networks are incompatible with the goals of this project. Instead, new models and algorithms will need to be developed, which are efficient in terms of the data and computational resources they require and transparent with respect to their inner workings.
SFB 1475, B01: Mixed-Methods Analysis of Medical Metaphorizations in Medieval German Texts
Reference Corpus of Middle High German
Reference Corpus of Early New High German
Roussel, Adam. 2023. “Lexical Semantics with Vector Symbolic Architectures.” In Proceedings of the Second Workshop on Resources and Representations for Under-Resourced Languages and Domains (RESOURCEFUL-2023), 53–61. Tórshavn, Faroe Islands. [PDF] [Repo]
Roussel, Adam, Fabian Barteld, and Katrin Ortmann. 2020. “CorA-XML Utils: Processing Diplomatic Transcriptions in Historical Corpora.” (Poster at DGfS 2020 in Hamburg, Germany.) [Repo]
Ortmann, Katrin, Adam Roussel, and Stefanie Dipper. 2019. “Evaluating off-the-Shelf NLP Tools for German.” In Proceedings of the 15th Conference on Natural Language Processing (KONVENS 2019): Long Papers, 212–222. Erlangen, Germany: German Society for Computational Linguistics & Language Technology. [PDF]
Dipper, Stefanie, Simone Schultz-Balluff, Marcel Bollmann, Julia Krasselt, Florian Petran, Adam Roussel, Katrin Ortmann, Katharina Bort, and Helena Wedig. 2018. “Anselm Corpus.” Ruhr-Universität Bochum: [Project Homepage]. Version 1.0. ISLRN 568-178-806-856-4.
Kolhatkar, Varada, Adam Roussel, Stefanie Dipper, and Heike Zinsmeister. 2018. “Anaphora With Non-Nominal Antecedents in Computational Linguistics: A Survey.” Computational Linguistics 44 (3): 547–612. [PDF].
Roussel, Adam. 2018. “Detecting and Resolving Shell Nouns in German.” In Proceedings of the NAACL Workshop on Computational Models of Reference, Anaphora, and Coreference (CRAC), 61–67. New Orleans, LA, USA. [PDF].
Roussel, Adam, Stefanie Dipper, Sarah Jablotschkin, and Heike Zinsmeister. 2018. “Towards the Automatic Resolution of Anaphora with Non-Nominal Antecedents: Insights from Annotation.” In Proceedings of the 14th Conference on Natural Language Processing (KONVENS). Vienna, Austria. [PDF].
Simonjetz, Fabian, and Adam Roussel. 2016. “Crosslinguistic Annotation of German and English Shell Noun Complexes.” In Proceedings of the 13th Conference on Natural Language Processing (KONVENS), 265–278. Bochum, Germany. [PDF].