Use of Adaptive Boosting Algorithm to Estimate User’s Trust in the Utilization of Virtual Assistant Systems

Akazue Maureen; Onovughe Anthonia; Edith Omede; Hampo, John Paul A.C.1

1

Publication Date: 2023/01/25

Abstract: User trust in technology is an essential factor for the usage of a system or machine. AI enabled technologies such as virtual digital assistants simplify a lot of process for humans starting from simple search to a more complex action like house automation and completion of some transitions notably Amazon’s Alexa. Can human actually trust these AI enabled technologies? Hence, this research applied adaptive boosting ensemble learning approach to predict users trust in virtual assistants. A technology trust dataset was obtained from figshare.com and engineered before training the adaptive boosting (AdaBoost) algorithm to learn the trends and pattern. The result of the study showed that AdaBoost had an accuracy of 94.31% for the testing set.

Keywords: Machine Learning, Ensemble Model, Predictive Model, Trust, Intelligent Virtual Assistants

DOI: https://doi.org/10.5281/zenodo.7568675

PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT23JAN727.pdf

REFERENCES

No References Available