Skripsi

PEMODELAN TOPIK MENGGUNAKAN PRE-TRAINED LANGUAGE MODEL ROBERTA DAN VARIATIONAL AUTOENCODER

---

Penilaian

---

The rapid and widespread flow of information highlights the importance of efficient text data management, making it even more important to organize and classify information from text data as more news is published online all the time. Topic modeling is useful in clustering news texts from the ever-growing sea of online news based on the topic of each text data. One method of topic modeling is to use Variational Autoencoder combined with a trained language model, RoBERTa. This research aims to create a topic modeling system using the Pre-trained Language Model RoBERTa and Variational Autoencoder. The dataset used consists of 5000 news data with 10 different topics taken from cnnindonesia, kompas, and detik.com. Topic modeling evaluation is done using coherence score cv, homogeneity score, and v-measure. With a coherence score cv of 77.3%, homogeneity score of 6.5%, and v-measure of 7.1%.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2407002531	T143437	T1434372024	Central Library (REFERENCES)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T1434372024

Publisher

Indralaya : Prodi Teknik Informatika, Fakultas Ilmu Komputer., 2024

Collation

xv, VI-2 hlm.; tab.; ilus.; 28 cm

Language

Indonesia

ISBN/ISSN

-

Classification

005.120 7

Content Type

Text

Media Type

unmediated

Carrier Type

unspecified

Edition

-

Subject(s)

Computer Software
Prodi Teknik Informatika

Specific Detail Info

-

Statement of Responsibility

UIN Farrah

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment