Consumers’ brand associations offer qualitative explanations on a brand’s success or failure and are typically elicited using survey-based instruments. Marketers are interested in time- and cost-efficient, automated brand association elicitation approaches. To enable an automated brand association elicitation, we show that brand associations can be formalized and described by patterns of linguistic part-of-speech sequences that differ from ordinary speech which is required for an automated extraction via text mining. Furthermore, we provide evidence that UGC is an adequate data-source for an automated brand association elicitation. We do that by comparing survey-based and UGC data-sources using linguistic part-of-speech sequence- and n-gram analysis as well as sequential pattern mining. We contribute to exiting research by establishing prerequisites for the construction of novel information systems that use text mining to extract brand associations automatically from UGC.
Volkmann, Gloria; Egger, Marc; and Schoder, Detlef, (2019). "TOWARDS MINING BRAND ASSOCIATIONS FROM USER-GENERATED CONTENT (UGC): EVIDENCE FROM LINGUISTIC CHARACTERISTICS". In Proceedings of the 27th European Conference on Information Systems (ECIS), Stockholm & Uppsala, Sweden, June 8-14, 2019. ISBN 978-1-7336325-0-8 Research Papers.