Hack your corpus analysis: How AI can assist corpus linguists deal with messy social media data
Michele Zappavigna
Abstract
This short reflection considers the role of Artificial Intelligence (AI) language models in supporting the work of corpus linguists. The focus is on how AI can be used in the analysis of complex and 'noisy' data commonly found in social media corpora. These datasets are characterized by unconventional language usage, shorthand, hashtags, emoji, and various other forms of imaginative expression. The discussion will be grounded in the author's ongoing research into emoji which offers an interesting example of the challenges of working with AI in the context of digital discourse analysis. The primary focus will be on how ChatGPT can support the creation and implementation of regular expressions and code for searching, matching, and manipulating text in corpora. The main objective is to guide the reader through the author's reasoning process as a way of reflecting on the integration of AI technology into the academic practice of corpus linguists.