Amrulloh, Dikri Fauzan (2026) IMPLEMENTASI WORD2VEC DAN COSINE SIMILARITY UNTUK PENILAIAN OTOMATIS JAWABAN ESAI PADA LMS MOODLE. S1 / D3 thesis, Universitas Kuningan.
ABSTRAK.pdf
Download (863kB)
BAB I.pdf
Download (208kB)
BAB II.pdf
Restricted to Repository staff only
Download (622kB) | Request a copy
BAB III.pdf
Restricted to Repository staff only
Download (482kB) | Request a copy
BAB IV.pdf
Restricted to Repository staff only
Download (1MB) | Request a copy
BAB V.pdf
Download (95kB)
DAFTAR PUSTAKA.pdf
Download (225kB)
LAMPIRAN.pdf
Restricted to Repository staff only
Download (1MB) | Request a copy
Abstract
Perkembangan teknologi pendidikan mendorong pemanfaatan Learning Management System (LMS) dalam proses evaluasi pembelajaran. Penilaian esai secara manual memiliki kendala karena membutuhkan waktu lama dan berpotensi menimbulkan subjektivitas. Penelitian ini bertujuan mengimplementasikan metode Word2Vec dan Cosine Similarity untuk penilaian otomatis jawaban esai pada LMS Moodle Universitas Kuningan. Sistem dikembangkan dalam bentuk question type plugin Moodle yang terintegrasi dengan REST API berbasis Python menggunakan metode Rapid Application Development (RAD). Proses penilaian menerapkan Natural Language Processing (NLP) melalui tahapan preprocessing teks, representasi Word2Vec, perhitungan Cosine Similarity, Keyword Coverage, serta penggabungan nilai menggunakan Weighted Score. Pengujian sistem dilakukan menggunakan White Box Testing, sedangkan evaluasi hasil penilaian menggunakan Mean Absolute Percentage Error (MAPE) dan Pearson Correlation. Hasil penelitian menunjukkan bahwa penambahan Keyword Coverage dan Weighted Score mampu menurunkan nilai MAPE dari 12,60% menjadi 5,72% serta meningkatkan Pearson Correlation dari 0,27 menjadi 0,92. Hasil tersebut menunjukkan bahwa sistem mampu menghasilkan penilaian otomatis yang mendekati penilaian dosen dan dapat digunakan sebagai alat bantu koreksi jawaban esai pada LMS Moodle.
The development of educational technology encourages the utilization of Learning Management Systems (LMS) in the learning evaluation process. Manual essay assessment requires considerable time and may lead to subjective evaluation. This study aims to implement Word2Vec and Cosine Similarity methods for automatic essay assessment in the Moodle LMS at Universitas Kuningan. The system was developed as a Moodle question type plugin integrated with a Python-based REST API using the Rapid Application Development (RAD) methodology. The assessment process applies Natural Language Processing (NLP) through text preprocessing, Word2Vec representation, Cosine Similarity calculation, Keyword Coverage measurement, and score combination using the Weighted Score method. System testing was conducted using White Box Testing, while assessment evaluation used Mean Absolute Percentage Error (MAPE) and Pearson Correlation. The results show that the addition of Keyword Coverage and Weighted Score reduced the MAPE value from 12.60% to 5.72% and increased the Pearson Correlation from 0.27 to 0.92. These results indicate that the system can generate automatic assessments that closely align with lecturer evaluations and can be used as a supporting tool for essay grading in Moodle LMS.
| Item Type: | Thesis (S1 / D3) |
|---|---|
| Uncontrolled Keywords: | Penilaian Otomatis Esai, Word2Vec, Keyword Coverage, Automatic Essay Scoring, Cosine Similarity, Weighted Score, Moodle |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science T Technology > T Technology (General) |
| Divisions: | Fakultas Ilmu Komputer > S1 Sistem Informasi |
| Depositing User: | S.Kom Dikri Fauzan Amrulloh |
| Date Deposited: | 14 Oct 2026 08:36 |
| Last Modified: | 14 Oct 2026 08:36 |
| URI: | https://rama.uniku.ac.id/id/eprint/6994 |

