Abstract
Personality is a unique trait that distinguishes an individual. It includes an ensemble of peculiarities on how people think, feel, and behave that affects the interactions and relationships of people. Personality is useful in diverse areas such as marketing, training, education, and human resource management. There are various approaches for personality recognition and different psychological models. Preceding work indicates that linguistic analysis is a promising way to recognize personality. In this work, a proposal for personality recognition relying on the dominance, influence, steadiness, and compliance (DISC) model and statistical methods for language analysis is presented. To build the model, a survey was conducted with 120 participants. The survey consisted in the completion of a personality test and handwritten paragraphs. The study resulted in a dataset that was used to train several machine learning algorithms. It was found that the AdaBoost classifier achieved the best results followed by Random Forest. In both cases a feature selection pre-process with Pearson’s Correlation was conducted. AdaBoost classifier obtained the average scores: accuracy = 0.782, precision = 0.795, recall = 0.782, F-measure = 0.786, receiver operating characteristic (ROC) area = 0.939.
Subject
Fluid Flow and Transfer Processes,Computer Science Applications,Process Chemistry and Technology,General Engineering,Instrumentation,General Materials Science
Cited by
6 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献