Skripsi

KLASIFIKASI TEKS KOMENTAR PRODUK PADA TOKOPEDIA MENGGUNAKAN METODE LONG SHORT-TERM MEMORY (LSTM)

---

Penilaian

---

Text data in comments is often unstructured, so to classify comments requires the application of appropriate methods. The method used in this study involves the use of algorithms from Deep Learning, namely Long Short-Term Memory (LSTM) to classify texts. In this study using 4060 imbalance dataset so that an Upsampling method is needed to adjust the distribution of data, so that the distribution of data changes to 8709, then it will be divided into 80% training data and 20% for testing data. The use of Word2Vec word embedding was also applied to this study. After tuning the LSTM hyperparameters, the final results were obtained using the adam optimizer, dropout layer of 0.5, hidden units of 400, 200, and 100 on each LSTM layer used, epochs of 100, batch size of 32 so that the accuracy obtained reached 80% with the average value of macro precision, macro recall, and macro F-measure of 36%, 36%, respectively. and 34% and accuracy obtained by 42%.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2307004839	T127117	T1271172023	Central Library (Referens)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T1271172023

Publisher

Indralaya : Jurusan Teknik Informatika, Fakultas Ilmu Komputer Universitas Sriwijaya., 2023

Collation

xv, 62 hlm.; Ilus.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

004.507

Content Type

-

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

penyimpanan
Jurusan Teknik Informatika

Specific Detail Info

-

Statement of Responsibility

SRI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment