A systematic mapping review of algorithms for the detection of rhymes, from early digital humanities projects to the rise of large language models

Abstract

We survey fifty years of algorithms to discover rhymes in natural language text, focusing largely on rhymes in English, but also in Italic and other Germanic languages. Using a systematic mapping review, we filtered from 4704 initially reviewed studies down to 89 that were relevant to our research questions and satisfied our inclusion criteria. Older papers document the history of simple computer algorithms being used to analyze poetry, but these also include some that create text with rhyming patterns. Papers from 2006 to 2016 often include complex algorithms for teasing out complex rhyme definitions, particularly in the domain of rap music. More recent papers have moved to studying the use of large language models (LLMs) and either adapting their mathematical properties, or simply training them on a collection of rhyming text. We explore how grey literature (blogs, open-source programming projects and more) relates to the academic literature in rhyme detection, and we describe the complexity of engaging in systematic reviews of this sort in areas that span many disciplines.