Share these talks and lectures with your colleagues
Invite colleaguesCase study
Topic modelling for open-ended survey responses
Applied Marketing Analytics: The Peer-Reviewed Journal,
4
(1),
53-62
(2018)
Abstract
Due to the availability of massive amounts of text data, both from online (Twitter, Facebook, online forums, etc) and offline open-ended survey questions, text analytics is growing in marketing research and analytics. Most companies are now using open-ended survey questions to solicit customer opinions on any number of topics (eg ‘how can we improve our service?’). With large sample sizes, however, the task of collating this information manually is practically impossible. This paper describes an end-to-end process to extract insight from text survey data via topic modelling. A case study from a Fortune 500 firm is used to illustrate the process.
Keywords: text analysis; open-ended questions; topic modelling; latent Dirichlet allocation; natural language processing
The full article is available to subscribers to the journal.
Author's Biography
Song Chen has a PhD in Applied Mathematics and is an assistant professor at University of Wisconsin – La Crosse. He conducts research in scientific computing and data mining with applications in fields such as marketing. He is director of the data science group at University of Wisconsin – La Crosse and actively leads undergraduate collaborative projects with local industry.
Chad Vidden has a PhD in Applied Mathematics, with expertise in computational mathematics, data science and machine learning. He is currently an assistant professor at the University of Wisconsin – La Crosse, where he leads a data science and mathematical modelling research group that collaborates with local companies.
Nicole Nelson is an analytical manager at Kwantum LLC, where she assists the Chief Data Scientist to conduct analytical projects and deliver the business solutions to clients. Nicole’s experience is specialised in key-driver modelling, marketing segmentation, Maxdiff analysis, data visualisation and text analytics. She has a chemistry degree and mathematics minor from University of Wisconsin — La Crosse.
Marco Vriens has a PhD in marketing analytics and is a recognised expert in applied analytics. He led analytics teams for Microsoft, GE and supplier firms. Marco is the author of three books: ‘The Insights Advantage: Knowing How to Win’ (2012), ‘Handbook of Marketing Research’ (2006) and ‘Conjoint Analysis in Marketing’ (1995). Marco has been published in academic and industry journals and has won several best paper awards including the David K. Hardin Award.
Citation
Chen, Song, Vidden, Chad, Nelson, Nicole and Vriens, Marco (2018, April 1). Topic modelling for open-ended survey responses. In the Applied Marketing Analytics: The Peer-Reviewed Journal, Volume 4, Issue 1. https://doi.org/10.69554/HOWE2138.
© Henry Stewart
Publications LLP
Publications LLP