IMPLEMENTASI ALGORITMA COSINE SIMILARITY UNTUK KLASIFIKASI JENIS PEKERJAAN DI BIDANG TEKNOLOGI INFORMASI
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
Wafit, Muhammad Istigfahri
Siswipraptini, Puji Catur
Djunaidi, Karina
Siswipraptini, Puji Catur
Djunaidi, Karina
Subject
Teknik Informatika
Datestamp
2023-07-17 07:13:14
Abstract :
Skills for each type of work must be prepared since students are in higher education. The type of work and its supporting skills become one of the foundations for the study program in formulating the specialization of courses according to the needs of the industry. So the study program needs guidelines to develop a curriculum that is in accordance with industry needs. This study aims to classify the types of work in the Information Technology (IT) field according to the 2020 Computing Curricula (CC2020 Task Force, 2020), issued by ACM as one of the guidelines for Higher Education (PT) in the field of Computer Science in Indonesia in compiling graduate competence and curriculum of PT. Job classification is carried out based on the needs of the industrial world as reflected in job vacancy advertisements and websites in Indonesia. The classification technique is carried out in several steps, namely web scraping on job vacancy websites; preprocessed text to clean web scraping data; and classification of 10 types of jobs using the Cosine Similarity algorithm. The cosine similarity algorithm analyzes the distance between documents for all types of work with required skills for job classification. The average cosine similarity is 0.242946529 or 24% and the highest similarity level is 0.63276049 or 63%. This test gets an accuracy value of 18.04% by using k-fold cross validation.
Skills for each type of work must be prepared since students are in higher education. The type of work and its supporting skills become one of the foundations for the study program in formulating the specialization of courses according to the needs of the industry. So the study program needs guidelines to develop a curriculum that is in accordance with industry needs. This study aims to classify the types of work in the Information Technology (IT) field according to the 2020 Computing Curricula (CC2020 Task Force, 2020), issued by ACM as one of the guidelines for Higher Education (PT) in the field of Computer Science in Indonesia in compiling graduate competence and curriculum of PT. Job classification is carried out based on the needs of the industrial world as reflected in job vacancy advertisements and websites in Indonesia. The classification technique is carried out in several steps, namely web scraping on job vacancy websites; preprocessed text to clean web scraping data; and classification of 10 types of jobs using the Cosine Similarity algorithm. The cosine similarity algorithm analyzes the distance between documents for all types of work with required skills for job classification. The average cosine similarity is 0.242946529 or 24% and the highest similarity level is 0.63276049 or 63%. This test gets an accuracy value of 18.04% by using k-fold cross validation.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
