Automatic Verb Extraction from Historical Swedish Texts

Even though historical texts reveal a lot of interesting information on culture and social structure in the past, information access is limited and in most cases the only way to find the information you are looking for is to manually go through large volumes of text, searching for interesting text segments. In this paper we will explore the idea of facilitating this timeconsuming manual effort, using existing natural language processing techniques. Attention is focused on automatically identifying verbs in early modern Swedish texts (1550–1800). The results indicate that it is possible to identify linguistic categories such as verbs in texts from this period with a high level of precision and recall, using morphological tools developed for present-day Swedish, if the text is normalised into a more modern spelling before the morphological tools are applied.