Responsive Banner

Klasifikasi pengaduan layanan pengguna Indihome pada media sosial twitter menggunakan metode support vector machine dengan seleksi fitur information gain

Ridho, Muhammad Ammarullah (2021) Klasifikasi pengaduan layanan pengguna Indihome pada media sosial twitter menggunakan metode support vector machine dengan seleksi fitur information gain. Undergraduate thesis, Universitas Islam Negeri Maulana Malik Ibrahim.

[img]
Preview
Text (FUlltext)
17650071.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (4MB) | Preview

Abstract

ENGLISH;

Social media is one of the media used to interact and communicate in informing others. One of the social media that is often used is Twitter. Many companies use Twitter as a customer service medium. One company that uses Twitter in this case, is PT. Telekomunikasi Indonesia (Telkom) to one of its products, namely IndiHome in response to his complaint. The complaints that users often write on Twitter are various service issues, bills, slow connections, and lost connections. In this study, the Support Vector Machine (SVM) method was used with the selection of the Information Gain feature. SVM is an appropriate method to classify these complaints into five classes. Meanwhile, Information Gain is used to reducing irrelevant features before the classification stage. The purpose of this study was to measure of accuracy, precision, recall, and f-measure in complaint classification. In the classification of complaints using the SVM and Information Gain features, it is known that the use of the best features is 80% which produces an accuracy value of 88.9%, precision 74.11%, recall 69.22% and f-measure 71.1% in service class (C1 ), 95% accuracy value, 87.2% precision, 87.8% recall and 87.4% f-measure in the installation class (C2), 94.4% accuracy value, 85.67% precision, 86.55 recall % and 86.07% f-measure in billing class (C3), 93% accuracy value, 82.47% precision, 82.62% recall and 82.48% f-measure in slow class (C4) and accuracy value. 89,1%, precision 71.93%, recall 74.48% and f-measure 73.07% in disconnect class (C5). Based on the use of features in the test scenario, it can be seen that the percentage of feature use in Information Gain affects the system classification results using the SVM method. So it can be concluded that the excessive selection of features will affect the classification results.

Item Type: Thesis (Undergraduate)
Supervisor: Hariri, Fajar Rohman and Suhartono, Suhartono
Contributors:
ContributionNameEmail
UNSPECIFIEDHariri, Fajar RohmanUNSPECIFIED
UNSPECIFIEDSuhartono, SuhartonoUNSPECIFIED
Keywords: Support Vector Machine; Information Gain; User Complaints
Subjects: 08 INFORMATION AND COMPUTING SCIENCES > 0801 Artificial Intelligence and Image Processing > 080107 Natural Language Processing
08 INFORMATION AND COMPUTING SCIENCES > 0801 Artificial Intelligence and Image Processing > 080109 Pattern Recognition and Data Mining
Departement: Fakultas Sains dan Teknologi > Jurusan Teknik Informatika
Depositing User: Muhammad Ammarullah Ridho
Date Deposited: 09 Dec 2021 09:58
Last Modified: 09 Dec 2021 09:58
URI: http://etheses.uin-malang.ac.id/id/eprint/29109

Downloads

Downloads per month over past year

Actions (login required)

View Item View Item