Pengenalan Emosi Pada Teks Baha Indonesia Dengan Metode @AM (Attitude Analysis Model)

Saepudin, Sidik (2017) Pengenalan Emosi Pada Teks Baha Indonesia Dengan Metode @AM (Attitude Analysis Model). Other thesis, Universitas Komputer Indonesia.

[thumbnail of UNIKOM_SIDIK SAEPUDIN_BAB 1.pdf] Text
UNIKOM_SIDIK SAEPUDIN_BAB 1.pdf

Download (141kB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_BAB 2.pdf] Text
UNIKOM_SIDIK SAEPUDIN_BAB 2.pdf

Download (1MB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_BAB 3.pdf] Text
UNIKOM_SIDIK SAEPUDIN_BAB 3.pdf

Download (2MB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_BAB 4.pdf] Text
UNIKOM_SIDIK SAEPUDIN_BAB 4.pdf

Download (2MB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_BAB 5.pdf] Text
UNIKOM_SIDIK SAEPUDIN_BAB 5.pdf

Download (70kB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_COVER.pdf] Text
UNIKOM_SIDIK SAEPUDIN_COVER.pdf

Download (17kB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_DAFTAR ISI.pdf] Text
UNIKOM_SIDIK SAEPUDIN_DAFTAR ISI.pdf

Download (106kB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_DAFTAR PUSTAKA.pdf] Text
UNIKOM_SIDIK SAEPUDIN_DAFTAR PUSTAKA.pdf

Download (221kB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_KATA PENGANTAR.pdf] Text
UNIKOM_SIDIK SAEPUDIN_KATA PENGANTAR.pdf

Download (93kB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_LEMBAR PENGESAHAN.pdf] Text
UNIKOM_SIDIK SAEPUDIN_LEMBAR PENGESAHAN.pdf

Download (508kB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_SURAT HAK EKSKLUSIF.pdf] Text
UNIKOM_SIDIK SAEPUDIN_SURAT HAK EKSKLUSIF.pdf

Download (1MB)
[thumbnail of UNIKOM_SIDIK SAEPUDIN_SURAT KETERANGAN TIDAK PLAGIAT.pdf] Text
UNIKOM_SIDIK SAEPUDIN_SURAT KETERANGAN TIDAK PLAGIAT.pdf

Download (454kB)

Abstract

Emosi yang tersembunyi di balik tulisan sangat sulit untuk ditafsirkan oleh pembaca dikarenakan pengaruh oleh sudut pandang yang berbeda antara penulis dan pembaca. Untuk dapat membantu mengenali emosi pada sebuah teks dapat menggunakan metode klasifikasi teks. Pada penelitian sebelumnya klasifikasi emosi dengan menggunakan metode KNN, dihasilkan nilai akurasi sebesar 50- 60%. Selain dengan metode perhitungan seperti KNN klasifikasi emosi dapat juga

dilakukan dengan pendekatan rules based, yang mana menurut hasil perbandingan antara metode klasifikasi dengan perhitungan dan rulses based diketahui bahwa pengklasifikasian teks dengan menggunakan rules based menghasilkan nilai akurasi yang lebih tinggi. Metode @AM (ATtitude Analysis Model) merupakan metode klasifikasi teks dengan pendekatan rules based. Rules atau aturan-aturan yang digunakan oleh metode @AM (ATtitude Analysis Model) diantaranya adalah polarity

reversal, aggregation, domination, netralization, dan intensification. Metode @AM (ATtitude Analysis Model) akan mengklasifikasikan teks ke dalam lima kategori yaitu �gembira�, �marah�, �jijik�, �sedih�, dan �takut�. Hasil perhitungan akurasi yang dihasilkan oleh sistem pengklasifikasian emosi dengan menggunakan metode @AM (ATtitude Analysis Model) pada data masukan tweet dari twitter sebanyak 500 buah tweet adalah sebesar 78,6%. Hasil akurasi tersebut sudah cukup baik, namun hasil tersebut sangat dipengaruhi oleh kualitas hasil survey yang dilakukan. Besaran akurasi survey sendiri dihitung menggunakan metode statistik fleiss kappa menghasilkan nilai kappa sebesar 0.267.

Emotions that are hidden behind the writing is very difficult to be interpreted by the reader due to influence by different viewpoints between the writer and the reader. To be able to help recognize emotion in a text can use text classification methods. Previous research on the classification of emotions using KNN, generated value accuracy by 50-60%. In addition to the method of calculation as KNN classification of emotions can also be performed with the approach of rules based, which, according to the results of a comparison between the method of classification by calculation and rulses based in mind that text classification using rules based produce higher accuracy values. @AM method (ATtitude Analysis Model) is a method of text classification with rules based approach. Rules that are used by the method @AM (ATtitude Analysis Model) include polarity reversal, aggregation, domination, netralization, and intensification. @AM method (ATtitude Analysis Model) will classify the text into five categories, namely "excited", "angry", "contempt", "sad", and "fear". The results of the calculation accuracy is produced by a system of classification of emotions by using the method @AM (ATtitude Analysis Model) on the data input tweet from twitter as much as 500 data tweet is 78.6%. The accuracy of results is already quite good, but the results are strongly influenced by the quality of the survey results is done. Magnitude accuracy survey itself is calculated using statistical methods fleiss kappa, kappa values is 0.267.

Item Type: Thesis (Other)
Additional Information: Migrasi dari GDL. Identifier: jbptunikompp-gdl-sidiksaepu-36793. Folder: S1-Final_Project / Fakultas_Teknik_dan_Ilmu_Komputer / Teknik_Informatika / 2017
Uncontrolled Keywords: penngenalan emosi, emosi, analisis sentimen
Divisions: S1 Skripsi > Teknik Informatika
Depositing User: Robot Migration
Date Deposited: 05 Nov 2026 12:31
Last Modified: 05 Nov 2026 12:31
URI: https://elibrary.unikom.ac.id/id/eprint/47790

Actions (login required)

View Item
View Item