Classification of TOGG Customer Complaints Using BERTopic Issue Modelling Technique

Özen, Üstün; Şahin, Veysel; Özenoğlu, YUNUS; Bölen, Mehmet

doi:10.21076/vizyoner.1676415

Classification of TOGG Customer Complaints Using BERTopic Issue Modelling Technique

Özen Ü., Şahin V., Özenoğlu Y. E., Bölen M. C.

SÜLEYMAN DEMIREL ÜNIVERSITESI VIZYONER DERGISI, cilt.17, sa.50, ss.404-425, 2026 (TRDizin)

Yayın Türü: Makale / Tam Makale
Cilt numarası: 17 Sayı: 50
Basım Tarihi: 2026
Doi Numarası: 10.21076/vizyoner.1676415
Dergi Adı: SÜLEYMAN DEMIREL ÜNIVERSITESI VIZYONER DERGISI
Derginin Tarandığı İndeksler: TR DİZİN (ULAKBİM)
Sayfa Sayıları: ss.404-425
Erzincan Binali Yıldırım Üniversitesi Adresli: Evet

Özet

In the study, TOGG’s 500,000 user messages are analyzed using the BERTopic topic modeling technique; the resulting data reveals that customer dissatisfaction is concentrated in financial (banking operations), technical (vehicle ınformation, software error, android compatibility), operational (online shopping), and service (delivery process, logistics problem) domains. The temporal variation of the complaints reflects the influence of periodic fluctuations and external factors, while the dendrogram analysis, by identifying the thematic proximities of topics, unveil the multidimensional nature of the customer experience. The findings emphasize the necessity for TOGG to address customer complaints through an integrated strategic planning process to create a competitive advantage. In the future, it is recommended to further enhance the application of BERTopic in Turkish by incorporating additional feedback from various data sources.