Peringkasan Otomatis Pada Multi Dokumen Menggunakan RVM

Arifin, Tendi (2017) Peringkasan Otomatis Pada Multi Dokumen Menggunakan RVM. Other thesis, Universitas Komputer Indonesia.

[thumbnail of 1. UNIKOM_TENDI ARIFIN_10113051_COVER.pdf] Text
1. UNIKOM_TENDI ARIFIN_10113051_COVER.pdf

Download (31kB)
[thumbnail of 4. UNIKOM_TENDI ARIFIN_10113051_KATA PENGANTAR.pdf] Text
4. UNIKOM_TENDI ARIFIN_10113051_KATA PENGANTAR.pdf

Download (193kB)
[thumbnail of 5. UNIKOM_TENDI ARIFIN_10113051_DAFTAR ISI.pdf] Text
5. UNIKOM_TENDI ARIFIN_10113051_DAFTAR ISI.pdf

Download (347kB)
[thumbnail of 10. UNIKOM_TENDI ARIFIN_10113051_DAFTAR PUSTAKA.pdf] Text
10. UNIKOM_TENDI ARIFIN_10113051_DAFTAR PUSTAKA.pdf

Download (264kB)
[thumbnail of 11. UNIKOM_TENDI ARIFIN_10113051_BAB I.pdf] Text
11. UNIKOM_TENDI ARIFIN_10113051_BAB I.pdf

Download (193kB)
[thumbnail of 12. UNIKOM_TENDI ARIFIN_10113051_BAB II.pdf] Text
12. UNIKOM_TENDI ARIFIN_10113051_BAB II.pdf

Download (667kB)
[thumbnail of 15. UNIKOM_TENDI ARIFIN_10113051_BAB V.pdf] Text
15. UNIKOM_TENDI ARIFIN_10113051_BAB V.pdf

Download (9kB)
[thumbnail of 18. UNIKOM_TENDI ARIFIN_10113051_LEMBAR PENGESAHAN.pdf] Text
18. UNIKOM_TENDI ARIFIN_10113051_LEMBAR PENGESAHAN.pdf

Download (343kB)
[thumbnail of 19. UNIKOM_TENDI ARIFIN_10113051_JURNAL BAHASA INDONESIA.pdf] Text
19. UNIKOM_TENDI ARIFIN_10113051_JURNAL BAHASA INDONESIA.pdf

Download (1MB)
[thumbnail of 20. UNIKOM_TENDI ARIFIN_10113051_JURNAL BAHASA INGGRIS.pdf] Text
20. UNIKOM_TENDI ARIFIN_10113051_JURNAL BAHASA INGGRIS.pdf

Download (889kB)
[thumbnail of 21. UNIKOM_TENDI ARIFIN_10113051_SURAT KETERANGAN TIDAK PLAGIAT.pdf] Text
21. UNIKOM_TENDI ARIFIN_10113051_SURAT KETERANGAN TIDAK PLAGIAT.pdf

Download (232kB)
[thumbnail of 22. UNIKOM_TENDI ARIFIN_10113051_SURAT KETERANGAN PERSETUJUAN PUBLIKASI.pdf] Text
22. UNIKOM_TENDI ARIFIN_10113051_SURAT KETERANGAN PERSETUJUAN PUBLIKASI.pdf

Download (231kB)

Abstract

Peringkasan teks otomatis pada multi dokumen adalah peringkasan banyak dokumen dengan tema yang sama menggunakan mesin. Penelitian sebelumnya, peringkasan teks otomatis pada multi dokumen dilakukan menggunakan metode Support Vector Machine dengan recall 0.773, precission 0.771, f-measure 0.771. Penelitian Christopher M. Bishop, Michael E. Tipping didapatkan tingkat error SVM lebih tinggi dari RVM yaitu SVM 10,6% dan RVM 9,3%. Tahapan dari peringkasan teks multi dokumen diantaranya penerimaan input teks, preprocessing, pembobotan ekstraksi fitur dengan fitur yang dipakai adalah posisi kalimat, kemiripan antar kalimat, kalimat yang menyerupai judul dokumen, kalimat yang mengandung kata entiti, kalimat yang mengandung data numerik, panjang kalimat, koneksi antar kalimat dan penjumlahan bobot koneksi antar kalimat kemudian pengolahan hasil ekstraksi fitur dengan RVM, pembobotan hasil RVM dengan tf-idf dan disusun menggunakan metode cosine similarity. Berdasarkan hasil pengujian disimpulkan bahwa Metode RVM dapat diterapkan pada Peringkasan Teks Otomatis Pada Multi Dokumen dengan performasi yang cukup kecil yaitu recall sebesar 47.83%, precission sebesar 35.48%, f-measure sebesar 40.74%, dan akurasi sebesar 67.35%.

Auto teks summarization on multi documents is teks summarization many documents with same theme using machine. On the research before, automatic teks summarization on multi documents is using Support Vector Machine methode and the result is recall 0.773, precission 0.771, f-measure 0.771. But on Christopher M. Bishop, Michael E. Tipping research have result Support Vector Machine have higher error than RVM (Relevance Vector Machine) that is SVM 10,6% and RVM 9,3%. Step on teks summarization is recieve document teks input, preprocessing, weighting proces with fitur ekstraktion and the fiturs is sentence postion, sentence similarity, similarity sentence with title of document, sentence have entity word, sentence have numeric data, long sentence, sentence connectivity, and count of sentence weight connection and next step is result of fiture extraction will be process using RVM, weighting again but on this step using tf-idf method and then the sentence will be stacking with cosine similarity method. Based on testing result automatix teks summariztion on multi documents using RVM method have performation result quite small the result is recall is 47.83%, precission is 35.48%, f-measure is 40.74%, dan akuration is 67.35%. it happens because the traing date is not balanced between important sentence and not important sentence.

Item Type: Thesis (Other)
Additional Information: Migrasi dari GDL. Identifier: jbptunikompp-gdl-tendiarifi-38464. Folder: S1-Final_Project / Fakultas_Teknik_dan_Ilmu_Komputer / Teknik_Informatika / 2017
Uncontrolled Keywords: Peringkasan Teks, peringkasan teks Otomatis pada muti dokumen, Relevance Vector Machine, Metode klasifikasi, machine learning RVM.
Divisions: S1 Skripsi > Teknik Informatika
Depositing User: Robot Migration
Date Deposited: 05 Nov 2026 13:00
Last Modified: 05 Nov 2026 13:00
URI: https://elibrary.unikom.ac.id/id/eprint/49353

Actions (login required)

View Item
View Item