SATRIA, Aditya Tegar and Mustafid, Mustafid and Mutiara K.N., Dinar (2020) SISTEM ANALISIS UMPAN BALIK PENUMPANG SINGAPORE AIRLINES PADA LAMAN TRIPADVISOR MENGGUNAKAN METODE RUMUS BAYES HIBRID DAN SUPPORT VECTOR MACHINE. Masters thesis, School of Postgraduate Studies.
Text
(COVER) Laporan Tesis_Aditya Tegar Satria_30000318410027 Ujian Tesis - Copy.pdf Download (1MB) |
|
Text
BAB 1 - Laporan Tesis_Aditya Tegar Satria_30000318410027 Ujian Tesis - Copy (2).pdf Download (425kB) |
|
Text
BAB 2 - Laporan Tesis_Aditya Tegar Satria_30000318410027 Ujian Tesis - Copy (3).pdf Download (695kB) |
|
Text
BAB 3 - Laporan Tesis_Aditya Tegar Satria_30000318410027 Ujian Tesis - Copy (4).pdf Restricted to Repository staff only Download (874kB) |
|
Text
BAB 4 - Laporan Tesis_Aditya Tegar Satria_30000318410027 Ujian Tesis - Copy (5).pdf Restricted to Repository staff only Download (1MB) |
|
Text
BAB 5 Laporan Tesis_Aditya Tegar Satria_30000318410027 Ujian Tesis - Copy.pdf Download (251kB) |
|
Text
DAFTAR PUSTAKA - Laporan Tesis_Aditya Tegar Satria_30000318410027 Ujian Tesis - Copy.pdf Download (318kB) |
Abstract
Algoritma Bayes dan Support Vector Machine merupakan metode yang umum digunakan untuk menyelesaikan permasalahan klasifikasi linier termasuk dalam penambangan data teks, namun pada umumnya kedua metode algoritma tersebut digunakan secara terpisah. Pada penelitian ini kedua metode algoritma tersebut digabungkan dalam satu sistem informasi untuk menganalisis kumpulan data teks dari ulasan penumpang pesawat Singapore Airlines pada laman TripAdvisor. Sistem yang dibangun ini bertujuan untuk klasifikasi kategori dan menganalisis aspek sentimen yang terkandung pada data ulasan penumpang dengan mengintegrasikan algoritma Bayes dan SVM pada modul inti pemrosesan dengan modul sistem lainnya seperti modul data masukan dan modul prapengolahan teks, serta mampu melakukan evaluasi sistem dengan data yang diujikan. Dari hasil proses terhadap 1000 dokumen yang diujikan, sistem menunjukkan tingkat akurasi algoritma Bayes mencapai 88,79% dan model algoritma SVM mencapai 84,76%. Penelitian ini telah mengimplementasikan metode baru dimana dua metode algoritma diterapkan secara hibrid yakni bekerja bersamaan pada satu sistem informasi dengan tetap mengoptimalkan kinerja masing-masing sehingga dapat menyajikan hasil yang lebih efektif, efisien, serta menunjukkan tingkat akurasi dan performa yang baik.
Kata kunci: penambangan teks, sistem hibrid, rumus Bayes, support vector machine
Bayes dan Support Vector Machine algorithms was commonly used to solve some linier classification problems such as text mining, but usually each of this method was particularly used as a single and independent method. On this research, both algorithms will be gathered and combine in a single information system to analyse large text dataset of Singapore Airline passengers review documents on TripAdvisor website. The main purpose of this information system is to perform a classification of document's categories and analysing sentiment aspect that contain on it by integrating Bayes and SVM algorithm method as main core module with other modules such as data input and text pre-processing, and also capable to perform system's evaluation. By processing 1000 amount of testing documents, system was able to show evaluation results of Bayes algorithm as 88,79% and SVM algorithm as 84,76%. This research has implemented a new method where two different algorithm methods has been combined and working simultaneously as a hybrid method on one single information system while optimizing each other's capabilities and performances so that it able to manage results that increase effectiveness, efficiency and show a good level of accuracy and performance.
Keywords: text mining, hybrid system, Bayes formula, support vector machine
Item Type: | Thesis (Masters) |
---|---|
Uncontrolled Keywords: | penambangan teks, sistem hibrid, rumus Bayes, support vector machine |
Subjects: | Sciences and Mathemathic |
Divisions: | Postgraduate Program > Master Program in Information System |
Depositing User: | ekana listianawati |
Date Deposited: | 26 Apr 2022 02:45 |
Last Modified: | 26 Apr 2022 02:45 |
URI: | https://eprints2.undip.ac.id/id/eprint/5970 |
Actions (login required)
View Item |