Automatic Personality Prediction: A Systematic Mapping Study

Context: Automatic Personality Prediction has experienced significant growth and has appealed to researchers thanks to its useful applications. However, as far as we are aware, there is a lack of studies that summarize systematically the different works in literature. - Objective: To describe the current state of the art, specifically to: (i) determine the demographic characteristics of the studies published in this topic, (ii) identify the used psychological models, (iii) characterize the prediction approaches in terms of methods, data and features. - Method: We conducted a systematic mapping study by following a recommended process. We employed an automatic search in different well-known databases and applied selection criteria to obtain the final set of papers to be studied. Then we proceeded to the extraction of specific information from selected papers to answer our questions. Results-: We evaluated 379 articles. The results show that the number of primary studies has increased significantly. We identified the trends concerning: personality models, data inputs and sources, prediction types, methods and features. - Conclusion: This overview of the literature may be a solid and comprehensive baseline for researchers in the domain of personality prediction. It may assist in identifying research directions and potential paths to explore.