Loading...
Paper Number
1974
Paper Type
Short
Description
Topic models have great potential for helping researchers and practitioners understand the electronic word of mouth (eWoM). This potential is thwarted by their purely unsupervised nature, which often leads to topics that are not entirely explainable. We develop a novel method to iteratively generate seed words to guide the interactive topic models. We assess the validity and applicability of the proposed method by investigating the critical phenomenon of Contact Tracing Mobile Applications (CTMAs) post-adoption during a time of the COVID-19 pandemic. The results show that constructs developed through our interactive topic modeling can capture primary research variables related to the phenomenon. Compared to existing topic modeling methods, our approach shows superior performance in explaining users’ satisfaction with CTMAs.
Recommended Citation
Namvar, Morteza; Akhlaghpour, Saeed; Boyce, James; and Sharifi Khajedehi, Salma, "Iterative Seed Word Generation for Interactive Topic Modelling: a Mixed Text Processing and Qualitative Content Analysis Approach" (2022). ICIS 2022 Proceedings. 6.
https://aisel.aisnet.org/icis2022/online_reviews/online_reviews/6
Iterative Seed Word Generation for Interactive Topic Modelling: a Mixed Text Processing and Qualitative Content Analysis Approach
Topic models have great potential for helping researchers and practitioners understand the electronic word of mouth (eWoM). This potential is thwarted by their purely unsupervised nature, which often leads to topics that are not entirely explainable. We develop a novel method to iteratively generate seed words to guide the interactive topic models. We assess the validity and applicability of the proposed method by investigating the critical phenomenon of Contact Tracing Mobile Applications (CTMAs) post-adoption during a time of the COVID-19 pandemic. The results show that constructs developed through our interactive topic modeling can capture primary research variables related to the phenomenon. Compared to existing topic modeling methods, our approach shows superior performance in explaining users’ satisfaction with CTMAs.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
20-OnlineReviews