Skripsi

ANALISIS SENTIMEN KOMENTAR POS ENDORSEMEN DI INSTAGRAM MENGGUNAKAN ALGORITMA SUPPORT VECTOR MACHINE DAN CHI-SQUARE. 

---

Penilaian

---

Instagram is the fourth most used social media with 1,38 billion active users worldwide. The rapid development of the number of Instagram users turns Instagram as one of the options to market products. Recently, Instagram celebrities emerge as they promote products and services of a business or widely known as endorsemen. Sentiment analysis is required to be done to determine the opinions of public or to discover the demeanor of consumers towards the marketed products or services. Support Vector Machine is one of the methods used to solve linier and non-linier problems. This method transforms every word that has been processed to features which turns the computation time longer. In order to solve the long computation time problem, a seleksi fitur feature is needed to reduce the number of features and improve the performance of SVM method. The objective of this research is to determine the performance of SVM method with Chi-Square seleksi fitur and the performance of SVM method without the seleksi fitur method in analysing the sentiments of endorsemen posts comments in Instagram. The results of the test will be compared to learn the impact of using feature selection method in order to improve the performance of SVM. The test is done using three SVM kernels, which are Linier, Polynomial, and RBF with C parameter values are 0,1, 1, and 10. The results of the test reveal that using seleksi fitur reduces the performance of SVM compared to using SVM method without seleksi fitur with polynomial kernel generates the best result. The drop of parameter values in confusion matrix are 3,6% in accuration, 18% in precision, 8% in recall, and 15% in f-measure.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2307000656	T89869	T898692022	Central Library (Referens)	Available

Detail Information

Series Title

-

Call Number

T898692022

Publisher

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

Collation

xvii, 139 hlm.; ilus.; 29 cm

Language

ISBN/ISSN

-

Classification

003.107

Content Type

Text

Media Type

unmediated

Carrier Type

-

Edition

-

Subject(s)

Prodi Teknik Informatika
Identifikasi Sistem

Specific Detail Info

-

Statement of Responsibility

MURZ

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment