AUTOMATIC RESUME SELECTION USING WORD SIMILARITY BASED ON WORDNET SCORE

• Main Author: Widyasari, Ratnadira
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/39799
• Institution: Institut Teknologi Bandung

One of the most prominent part of an organization or company is human resources. Quality human resources is a necessity for a good company. Acquiring good human resources starts from recruitment and selection process. Selection process is used to decide the best candidate to fill certain position or placement. This process can be difficult for a company due to the complexities of the criteria and the number of candidates. Resume matching system which are built use a rule-based technique. Rule-based technique is used because the input data from the resume is structured and easy to process. Moreover, there was a difficulty on collecting a big amount of resume and criteria data to make a statistical-based system. Therefore, it is decided that rule-based system is the most suitable technique for this case. The system take advantage of wordnet to match between candidate’s resume and the given criteria. For each criterion, the corresponding resume input is scored and sorted based on the highest score. Higher score will mean that the candidate is more suitable for the job. The result of the system functionality test is consistent with the actual test result. Furthermore, resume matching for each criterion achieve an accuracy of 97%. Adjustment of scoring weight during resume matching results in a 100% accuracy for resume matching. Usability testing is conducted on job applicants and human resource department members with SUS score of 85.9 and 88.1. It can be concluded that the usability of the system is high and the system is easy to use.