<![CDATA[This study analyzes the compatibility between competencies in the National Occupation Map (SKKNI) of the ICT sector under the Web Developer scheme and competency requirements found in online job vacancies in Indonesia. The study employs Natural Language Processing (NLP) with Multilingual BERT (mBERT) to generate semantic representations of competency-related texts and Cosine Similarity to measure their semantic similarity. The dataset consists of 18 competency units, 58 competency elements, and 157 Performance Criteria (Kriteria Unjuk Kerja/KUK) from the National Occupation Map, along with 255 online job vacancies collected from Jobstreet Indonesia and LinkedIn. In the similarity analysis, all Performance Criteria and Competency Elements were aggregated into separate text representations and compared with the text representation of each job vacancy. The results show a mean similarity of 0.792932 for the aggregated KUK and 0.800890 for the aggregated Competency Elements, categorized as Suitable and Highly Suitable, respectively. The similarity values ranged from 0.646282 to 0.940792 for the aggregated KUK and from 0.639456 to 0.947454 for the aggregated Competency Elements. Word occurrence analysis further identified C, API, and Java as the three most frequently required competencies, appearing in 65.10%, 48.24%, and 45.10% of the analyzed vacancies, respectively. These results indicate that the competencies represented in the National Occupation Map generally show a high level of semantic compatibility with competency requirements in online job vacancies.]]>
Copyrights © 2026
