A machine learning approach to employability evaluation using handwriting analysis

Abstract

Writing is a process which is rooted deep within the creative processes of the mind. While writing, the mind and the hand synchronize to act as a smoothly functioning unit, thus enabling the pen to act as an extension of the person’s innermost self. Emotional factors often dictate the writing strokes and mannerisms. The science of graphology can be applied to discern a person’s behavior and inner psychological makeup from their handwriting. Certain features such as the page margins, handwriting size etc. are often reflective of mood changes and characterize the writer’s state of mind at the moment of writing. An automated process for extracting these features and mapping them to the various personality traits can definitely prove to be a boon for many applications like - recruitment process or even psychological analysis. In this paper, we propose feature extraction methods implemented using image processing techniques to select features to be used further for this trait identification. Once the features have been selected, existing classifiers have been put to work to determine the employability evaluation of a candidate from an HR perspective.