Mining brand-related tweets for brand monitoring with consumer-based brand equity classification and sentiment analysis

The growth of social media makes it possible for brand managers to monitor consumers’ perceptions and attitudes towards brand in a new way, which is timelier and more cost-efficient. The consumers’ responses to brand are also known as consumer-based brand equity (CBBE). Guided by brand equity literature, this dissertation provides a framework of how to analyze consumer-generated and brand-related textual data on Twitter effectively and insightfully with two studies, which involves qualitative content analysis and machine learning modelling. Study 1 focused on the development of the CBBE classification scheme. Along with brand equity literature, unsupervised topic modelling is conducted to aid the qualitative content discovery. Two coders then coded a sample of tweets as training data into one of the CBBE dimensions and meanwhile, they also labelled sentiment for the tweets. Study 2 utilized machine learning to help label data and answer research questions. Different natural language processing techniques and models are compared and summarized. It is found that the models with the Bidirectional Encoder Representations from Transformers (BERT) embedding technique applied have the best performance. Methodological implications are discussed. Besides, several use cases are provided in Study 2 to illustrate how the CBBE classification and sentiment analysis can be used together to generate consumer insights. Practical implications to brand and advertising are also discussed.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Communications and Media

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2022

Author dc:creator

• Yao, Jiachen

Contributors dc:contributor

• Sar, Sela
• Yang, Feng
• Yao, Mike
• Su, Leona (Yi-Fan)
• Ham, Chang-Dae

• social media
• brand equity
• data analytics
• BERT

Statement dc:rights

• Copyright 2022 Jiachen Yao

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/116247

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F116247