BibTeX

@techreport{morger-2024-when-342179,
	title        = {When Sparv met Superlim. . . A Sparv Plugin for Natural Language Understanding Analysis of Swedish},
	abstract     = {This technical report introduces Sparv-Superlim, a Sparv plugin for natural language understanding analysis of Swedish. It uses the reference models trained on the Superlim multi-task benchmark to add additional analyses to the Sparv Pipeline. I show how to install and configure the tool as well as apply it to analyze Swedish political manifestos to see if the predictions the plugin does align with known political positions of Swedish parties. These use cases shows that the reference models vary in their applicability to predict correct sentiments on novel data and illustrates the importance of integrating reference models trained on a multi-task benchmark like Superlim to evaluate the ecological validity of the benchmark.},
	author       = {Morger, Felix},
	year         = {2024},
}