TY - GEN
T1 - Auto-tracking controversial topics in social-media-based customer dialog
T2 - 13th International Conference on Transforming Digital Worlds, iConference 2018
AU - Yu, Bei
AU - Yu, Yihan
N1 - Publisher Copyright:
© Springer International Publishing AG, part of Springer Nature 2018.
PY - 2018
Y1 - 2018
N2 - This study proposed and validated a topic modeling-based approach for auto-tracking customer dialog on social media, using Starbucks as a case study because of its pioneering social media practice in service industry. A topic model was fit based on nearly 150,000 customer comments posted to Starbucks’ Facebook page in 2013. This model was able to identify not only business-related topics, such as customer responses to marketing campaigns, but also controversial topics regarding community involvement and corporate social responsibility, such as gay, gun, and government. Guided by this topic model, each topic’s evolving dynamics and patterns of user participation were further revealed, providing a bird’s-eye view of the topics and their evolution. The case study has demonstrated that the proposed approach can effectively track the main themes in the customer dialog on social media, zoom in on the controversial topics, measure their time spans, and locate the participants and the vocal activists. Such information would be valuable input for companies to design their intervention strategies and evaluate the outcomes in social media discussions.
AB - This study proposed and validated a topic modeling-based approach for auto-tracking customer dialog on social media, using Starbucks as a case study because of its pioneering social media practice in service industry. A topic model was fit based on nearly 150,000 customer comments posted to Starbucks’ Facebook page in 2013. This model was able to identify not only business-related topics, such as customer responses to marketing campaigns, but also controversial topics regarding community involvement and corporate social responsibility, such as gay, gun, and government. Guided by this topic model, each topic’s evolving dynamics and patterns of user participation were further revealed, providing a bird’s-eye view of the topics and their evolution. The case study has demonstrated that the proposed approach can effectively track the main themes in the customer dialog on social media, zoom in on the controversial topics, measure their time spans, and locate the participants and the vocal activists. Such information would be valuable input for companies to design their intervention strategies and evaluate the outcomes in social media discussions.
KW - Customer dialog
KW - Social media
KW - Text mining
KW - Topic modeling
UR - http://www.scopus.com/inward/record.url?scp=85044410435&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85044410435&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-78105-1_11
DO - 10.1007/978-3-319-78105-1_11
M3 - Conference contribution
AN - SCOPUS:85044410435
SN - 9783319781044
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 87
EP - 96
BT - Transforming Digital Worlds - 13th International Conference, iConference 2018, Proceedings
A2 - Chowdhury, Gobinda
A2 - McLeod, Julie
A2 - Gillet, Val
A2 - Willett, Peter
PB - Springer Verlag
Y2 - 25 March 2018 through 28 March 2018
ER -