Pengklasifikasian Topik dan Analisis Sentimen Dalam Media Sosial

Susilo, Tegar Heru and Rochimah, Siti (2013) Pengklasifikasian Topik dan Analisis Sentimen Dalam Media Sosial. In: Seminar Nasional Sistem & Teknologi Informasi (SNASTI) 2013, 24 Oktober 2013, Surabaya.

Text 2013-SC-01.pdf - Published Version Download (566kB)

Search this title on : |

Abstract

Social media has the capability to increase student’s potencies (Sturgeon, 2009) measured by intellectual, social, and performance level. This potency is affected by informal relation between lecturer and students. In the other hand, lecturer has the responsibilities for teaching, research, and public service (Peraturan Pemerintah no 37 tahun 2009). In relation to teaching, lecturer is responsible for guiding students. Guidance has meant to direct students to learn and have a good behavior by encouraging and providing examples. Therefore, lecturer should know, and understand the emotions his students have in order to provide appropriate treatment. These emotions can be seen from the student’s statuses in social media. In this research, an application is proposed. This application has the ability to retrieve information about student statuses in social media, doing topic classification using SVM (Yu, 2011) between academic and non-academic label, and doing sentiment analysis using Maximum Entropy (Soria, 2010) between positive and negative emotions. Testing was conducted in a form of dataset testing using learn and classify approach for testing SVM and MaxEnt classification result. In the dataset testing for SVM, the result shows an accuracy rate of 93%. While in the dataset testing for MaxEnt, the result shows an accuracy rate of 70% for positive document and 53% for negative document. Improved accuracy of sentiment analysis is obtained from the use of word-shape feature in the learning process.

Export Record

ASCII Citation Atom BibTeX Dublin Core EP3 XML EndNote HTML Citation JSON METS MODS MPEG-21 DIDL OpenURL ContextObject OpenURL ContextObject in Span RDF+N-Triples RDF+N3 RDF+XML Refer Reference Manager Simple Metadata

Download Statistics

Downloads over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Actions (login required)

View Item