Seamless Transition: Mapping Smartphone-Derived Personality Profiles to Business Intelligence

This study explores the feasibility of transferring a personality prediction model, trained onsmartphone data, to a Business Intelligence (BI) system utilizing call logs. By leveraging theBig Five personality traits model and machine learning techniques, the research addresses thechallenge of harmonizingdisparate data sources—richdataaboutsmartphoneapplicationsusageand system settings, and aggregated call log data from a BI warehouse. Through meticulous datapreprocessing and feature engineering, the study evaluates the performance of various machinelearning algorithms, demonstrating that traits like Extraversion and Conscientiousness can bereliably predicted. The study also introduces innovative methods for verifying predictions onunlabeled BI data, including consistency over time and similarity of distributions. The findingsunderscore the potential of integrating personality insights into BI systems, also highlighting theethical considerations and data privacy challenges inherent in such applications. This researchcontribution lays the groundwork for future advancements in personality prediction models andtheir practical business applications.