A Talent Assessment Model Based on Learning Behaviors and Patterns

Abstract

Talent assessment is an important topic in various areas like enterprise management, education, and psychology. However, it is also a challenging topic as the conventional assessment methods and models are unsuitable for talent assessment due to the following two aspects: (i) domain-dependent. The assessment of talent is highly depended on a specific domain which requires a large volume of domain knowledge in the assessment model; and (ii) behavior-based (or pattern-based). The characteristics of talents are reflected by a wide range of factors like their behaviors (patterns), emotions, self-identities, and metacognition. In this paper, we propose a talent assessment model based on online learning behaviors and patterns by using fuzzy models. Specifically, we attempt to develop a talent assessment model by identifying their learning data as we believe that the learning behaviors in the online learning platforms like massive open online courses (MOOCs) can reflect some characteristics of talents. Furthermore, we discuss what are the data sources, the learning behaviors and the potential computational methods in this assessment model in details. In addition, the potential limitations and possible improvement plans are introduced.