To identify and explore the linguistic properties of Norwegian constructions, EliNor will need to address the data representative of informal Norwegian. The existing Norwegian corpora have limitations. There is a need for a new corpus that is more “focused” (Barbaresi 2019) than the Norwegian web corpus NoWaC, includes sociodemographic information about language users, contains language that is more informal and spoken-like than that of Leksikografisk bokmålskorpus, and is significantly larger and more recent than the available spoken corpora (Nordisk dialektkorpus, the BigBrother-corpus).

To fill this gap, EliNor builds the Norwegian Blog Corpus that will serve as the major data source for the Norwegian Constructicon. The corpus is compiled by Olaf Mikkelsen (https://github.com/omikke/NBC) and contains 18 million words from recent blog texts written by more than 800 bloggers in 2010-2022. The Corpus requires further work on data cleaning and additional annotations. Currently, it is at Technology Readiness Level 3. EliNor aims to advance this resource to Technology Readiness Levels 6-7 by developing a modern, professional interface with advanced filtering options, including sociodemographic variables, time periods, and stylistic features. This will ensure that the Corpus becomes a versatile and indispensable resource for linguistic research and the development of Norwegian language technologies.