A text mining and topic modelling perspective of ethnic marketing research

Abstract

This study presents an enhanced automated approach based on literature analysis and synthesis for establishing the dimensions of the ethnic marketing literature, covering a set of 239 journal articles published by nine major publishers. The approach reported is enhanced by two novel procedures to address previously identified limitations, namely: definition of a relevant dictionary based on both a sufficient lexicon extracted from a definition of the core theme and a conditional dictionary, with related but non-core terms; and a visually appealing pictorial representation to summarize the discovered topics. The application of the method to ethnic marketing indicates that ethnic marketing research is characterized by high conceptual heterogeneity, although a clear definition of "ethnic marketing" is imperative for research development. Overall, the paper advances an approach with considerable scalability advantages when compared with extant approaches, an important issue to consider when textual sources become big data.