PENERAPAN METODE VECTOR SPACE MODEL (VSM) DENGAN TF-IDF DAN COSINE SIMILARITY PADA SISTEM TEMU KEMBALI INFORMASI LOWONGAN PEKERJAAN

Abstract

Finding relevant job vacancy information is often a challenge for job seekers due to the large volume of data on the internet and the limitations of exact keyword matching, which often produces inaccurate results. This study aims to build a job vacancy Information Retrieval (IR) system using the Vector Space Model (VSM) approach. The system applies Term Frequency-Inverse Document Frequency (TF-IDF) for term weighting and Cosine Similarity to measure the relevance between user queries and job documents. The study utilized a dataset of 350 documents, consisting of 300 valid job postings and 50 noise documents. The text preprocessing stages included case folding, tokenizing, filtering, and stemming. System performance was evaluated using Precision, Recall, and F-Measure metrics on 10 different search queries. The test results demonstrated high accuracy in retrieving relevant documents, achieving an average Precision of 0.840 (84%) at the top-10 threshold and 0.880 (88%) at the top-15 threshold. These results indicate that the combination of TF-IDF and Cosine Similarity is effective in filtering out irrelevant documents and ranking job vacancies according to the user's needs.