Assessment of Non-Verbal Communication Online Job Recruitment Using Gray Level Co-Occurrence Matrix (GLCM) and Fuzzy C-Means Algorithm (FCM)
Anita Sindar Sinaga

STMIK Pelita Nusantara

Abstract

Online recruitment, interviewers are more focused on the applicant^s body language. Through video recordings from Skype, video calls anyone can apply. An online recruitment system can be used to analyze body language. Communication contains two dimensions, verbal and non-verbal. Non-verbal communication behavior was evaluated using the feature extract calculation of the Gray Level Co-Occurrence Matrix (GLCM). Body language assessed by eyes, facial expressions, intonation, volume of voice and physical appearance. Applicants^ videos are extracted to take eye, mouth and head gestures. Grayscale gestures were converted using the Local Binary Pattern (LBP) method. Image analysis based on order 1 and order2 statistical distribution using GLCM. The feature extraction formula consists of feature, contrast, energy, entropy and homogeneity at rotation 00, 450, 900 and 1350. The GLCM method is an effective texture descriptor method and has better computation time and accuracy than other texture extraction methods. The results of the GLCM extraction calculation are then classified using the Fuzzy C-Means algorithm. Fuzzy C-Means (FCM) is used in pattern recognition. Data identification on the FCM is determined based on the degree of membership which has a value between 0 and 1. This study processes the applicant^s offline video data. Identify expressions that indicate doubt, confusion and optimism. The assessment of the 10 videos was taken with the dominant optimistic expression extraction loop, it was obtained that the 7th video had the highest percentage of optimistic facial expressions.

Keywords: Offline Video, Non Verbal Communication, LBP, FCM, GLCM

Topic: Engineering