Universal natural language understanding (completed)

• Salimifar, Saeedeh (2024). Presupposition projection: let’s test again! Proceedings of ConSOLE. ISSN 1574-499X. 31, p. 93–108. Full text in Research Archive
• Torgersen, Henrik Anker Tumert; Ravnanger, Erlend Øyvindssønn; Hellan, Lars & Haug, Dag Trygve Truslew (2024). NoVRol: A semantic role lexicon of Norwegian verbs. In Calzolari, Nicoletta; Kan, Min-Yen; Hoste, Veronique; Lenci, Alessandro; Sakti, Sakriani & Xue, Nianwen (Ed.), Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). European Language Resources Association. ISSN 9782493814104. Full text in Research Archive Show summary

  In this paper, we describe NoVRol, a semantic role lexicon of Norwegian verbs. We start from the NorVal valency lexicon, which describes the syntactic frames of 7.400 verbs. We then enrich each of these frames by annotating, based on the VerbNet annotation scheme, each argument of the verb with the semantic role that it gets. We also encode the syntactic roles of the arguments based on the UD annotation scheme. Our resource will faciliate future research on Norwegian verbs, and can at a future stage be expanded to a full VerbNet

• Haug, Dag Trygve Truslew (2023). LFG and Dependency Grammar. In Dalrymple, Mary (Eds.), The Handbook of Lexical Functional Grammar. Language Science Press. ISSN 9783985540822. p. 1829–1859. doi: 10.5281/zenodo.10186040. Full text in Research Archive
• Haug, Dag Trygve Truslew; Yildirim, Ahmet; Hagen, Kristin & Nøklestad, Anders (2023). Rules and neural nets for morphological tagging of Norwegian - Results and challenges. In Alumäe, Tanel & Fishel, Mark (Ed.), Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa). University of Tartu. ISSN 9789916219997. p. 425–435. Full text in Research Archive Show summary

  This paper reports on efforts to improve the Oslo-Bergen Tagger for Norwegian morphological tagging. We train two deep neural network-based taggers using the recently introduced Norwegian pre-trained encoder (a BERT model for Norwegian). The first network is a sequence-to-sequence encoder-decoder and the second is a sequence classifier. We test both these configurations in a hybrid system where they combine with the existing rule-based system, and on their own. The sequence-to-sequence system performs better in the hybrid configuration, but the classifier system performs so well that combining it with the rules is actually slightly detrimental to performance.

• Findlay, Jamie Yates; Yildirim, Ahmet; Salimifar, Saeedeh & Haug, Dag Trygve Truslew (2023). Rule-based semantic interpretation for Universal Dependencies. In Grobol, Loïc & Tyers, Francis (Ed.), Proceedings of the Sixth Workshop on Universal Dependencies. Association for Computational Linguistics (ACL). ISSN 9781959429340. p. 47–57. Full text in Research Archive
• Haug, Dag Trygve Truslew; Findlay, Jamie Yates & Yildirim, Ahmet (2023). The long and the short of it: DRASTIC, a semantically annotated dataset containing sentences of more natural length. In Bonn, Julia & Xue, Nianwen (Ed.), Proceedings of the Fourth International Workshop on Designing Meaning Representations. Association for Computational Linguistics (ACL). ISSN 9781959429654. p. 89–98. Full text in Research Archive
• Findlay, Jamie Yates & Haug, Dag Trygve Truslew (2022). Managing scope ambiguities in Glue via multistage proving. Proceedings of the LFG-conference. p. 144–163. Full text in Research Archive Show summary

  It is well known that vanilla LFG+Glue overgenerates when it comes to scope ambiguities. This has both theoretical and practical implications: it generates unattested readings, and also produces many spurious ambiguities which do not correspond to linguistically relevant differences. We propose a solution in keeping with the LFG philosophy: a new level of the parallel projection architecture, called proof-structure, which can be used to limit scopal interactions. Unlike earlier attempts to solve this problem, the modularity of our approach means that we do not need to alter the underlying linear logic or proof algorithm of Glue Semantics, with the effect that existing analyses and tools continue to be relevant and usable.

• Findlay, Jamie Yates & Haug, Dag Trygve Truslew (2021). How useful are Enhanced Universal Dependencies for semantic interpretation? In Osenova, Petya (Eds.), Proceedings of the Sixth International Conference on Dependency Linguistics (Depling, SyntaxFest 2021). Association for Computational Linguistics (ACL). ISSN 9781955917148. p. 22–34. Full text in Research Archive Show summary

  We discuss the role of enhanced Universal Dependencies (E-UD) in the task of deriving semantic predicate-argument structures from UD treebanks in a universal, non-language-specific way. We consider the usefulness of three kinds of E-UD annotation (controllers of xcomps, propagation of outgoing dependencies in coordinations, and coreference in relative clauses) and assess some heuristics for automatically adding such enhancements. We conclude that one large obstacle both for deriving predicate-argument structures from UD treebanks and for the automatic enhancement of basic UD treebanks is the fact that UD does not represent empty elements such as pro-dropped arguments, and we suggest that devoting effort to this would often provide a better return on investment than spending resources on improving or adding E-UD annotations.

• Findlay, Jamie Yates (2021). Meaning in LFG. In Arka, I Wayan; Asudeh, Ash & King, Tracy Holloway (Ed.), Modular Design of Grammar: Linguistics on the Edge. Oxford University Press. ISSN 9780192844842. p. 340–374. Full text in Research Archive

View all works in NVA