News discourse about migration to the UK (2011-2016) surrounding Brexit: recurring themes and their relationship to migration flows

This work investigated the depiction of migrants in the UK by a right-wing, national newspaper, The Daily Mail. Computational, Machine Learning techniques were used to extract recurrent themes and examine the connection between depictions and empirical net immigration levels. The analysis was performed on a corpus of news articles recovered from the LexisNexis archive, published between 1st of January 2011 and the 31st December 2016, using migrant-related keywords. Recurrent themes were automatically extracted using BERTopic. Results were compared to estimates for net immigration during the time period. The relationship between frequency of articles, themes present, evolution of theme representations and net migrations are discussed. Topics identified were found to be broadly consistent with existing themes described in the literature. However, themes of cultural integration were absent from the findings, while a topic relating to sport and athletes was identified that was not previously described by research. A moderate link was found between article frequency and empirical migration estimates.