Chapter 5. Semi-automatic approaches to Anglicism detection in Norwegian corpus data

This article describes corpus-based research methods and language processing tools that are used for the systematic study of the influence of English on Norwegian lexis. The tools are developed in connection with the Norwegian Newspaper Corpus (NNC) project. The study presents a survey of the types of phenomena that an Anglicism detection tool should aim at identifying and the problems associated with the orthographic and morphological variability of Anglicisms. It also describes the development of an Anglicism detection tool and accounts for a set of experiments using lexicon-based, n-gram-based and combinatory methods. Finally it describes recently developed machine learning techniques that have been developed by the NNC team, arguing that the computational approach to Anglicism identification is a fruitful one.