RIYANTO, BIMO BAGAS (2026) PEMETAAN TREN PENELITIAN AKADEMIK BERDASARKAN JUDUL PUBLIKASI MENGGUNAKAN WORD2VEC DAN K-MEANS CLUSTERING. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
Cover.pdf Download (412kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
Bab 1.pdf Restricted to Registered users only Download (95kB) |
|
|
Text (BAB II)
Bab 2.pdf Restricted to Registered users only Download (140kB) |
|
|
Text (BAB III)
Bab 3.pdf Restricted to Registered users only Download (83kB) |
|
|
Text (BAB IV)
Bab 4.pdf Restricted to Registered users only Download (485kB) |
|
|
Text (BAB V)
Bab 5.pdf Restricted to Registered users only Download (27kB) |
|
|
Text (DAFTAR PUSTAKA)
Daftar Pustaka.pdf Restricted to Registered users only Download (91kB) |
|
|
Text (LAMPIRAN)
Lampiran.pdf Restricted to Registered users only Download (631kB) |
Abstract
Academic institutions face increasing challenges in systematically mapping the distribution and trends of faculty research topics. This study proposes an approach that combines Word2Vec word embedding (CBOW architecture, 100 dimensions) with K-Means Clustering for automatic topic modeling of 2,107 publication titles from lecturers at the Faculty of Computer Science, University Y, collected through automated extraction from Google Scholar. This approach was chosen for its ability to capture semantic relationships between academic terms, in contrast to TF-IDF, which relies solely on lexical frequency. The optimal number of clusters (k=4) was determined using the Elbow Method and validated through the Silhouette Score, yielding a value of 0.3022 from 1,946 unique titles distributed across four clusters (1,068, 637, 223, and 18 documents), with dominant topics including application and data-based model development, information system and decision support system design, machine learning applications for classification and prediction, and cybersecurity and parallel systems. The Information Systems study program showed greater dominance in system design topics, while Informatics Engineering was more dominant in machine learning applications. Trend analysis from 1983 to 2026 revealed a sharp turning point in growth between 2017 and 2019, with the application development and information systems clusters consistently dominating, the machine learning cluster showing significant growth since 2018, and the cybersecurity cluster remaining sparsely represented throughout the observation period. Keywords: research trend mapping; word2vec; k-means clustering; natural language processing; publication title analysis Institusi akademik menghadapi tantangan dalam memetakan distribusi dan tren topik penelitian dosen secara sistematis. Penelitian ini mengusulkan pendekatan yang menggabungkan word embedding Word2Vec (arsitektur CBOW, dimensi 100) dengan K-Means Clustering untuk pemodelan topik otomatis terhadap 2.107 judul publikasi dosen Fakultas Ilmu Komputer, Universitas Y, dikumpulkan melalui ekstraksi otomatis dari Google Scholar. Pendekatan ini dipilih karena mampu menangkap hubungan semantik antar istilah akademik, berbeda dengan TF�IDF yang hanya mengandalkan frekuensi leksikal. Jumlah klaster optimal (k=4) ditentukan menggunakan Metode Elbow dan divalidasi melalui Silhouette Score, menghasilkan nilai 0,3022 dari 1.946 judul unik yang terbagi ke dalam empat cluster (1.068, 637, 223, dan 18 dokumen) dengan topik dominan: pengembangan aplikasi dan model berbasis data, perancangan sistem informasi dan sistem pendukung keputusan, penerapan machine learning untuk klasifikasi dan prediksi, serta cybersecurity dan sistem paralel. Program Studi Sistem Informasi lebih dominan pada topik perancangan sistem, sementara Teknik Informatika lebih dominan pada machine learning. Analisis tren tahun 1983–2026 menunjukkan titik balik pertumbuhan pesat pada 2017–2019, dengan cluster aplikasi dan sistem informasi konsisten mendominasi, cluster machine learning tumbuh signifikan sejak 2018, dan cluster cybersecurity tetap jarang muncul sepanjang periode pengamatan. Kata Kunci: pemetaan tren penelitian; word2vec; k-means clustering; natural language processing; analisis judul publikasi
Actions (login required)
 |
View Item |
