Article Per Year (5 Year)
p-Index From 2020 - 2025
0.23
P-Index
This Author published in this journals
All Journal Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Fira Sukmanisa
Universitas Brawijaya
Computer Science & IT Control & Systems Engineering Education Electrical & Electronics Engineering Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Deteksi Emosi pada Tweet Berbahasa Indonesia tentang Pembelajaran Jarak Jauh Menggunakan K-Nearest Neighbor dengan Pembobotan Kata Term Frequency-Inverse Gravity Moment Fira Sukmanisa; Yuita Arum Sari; Imam Cholissodin
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 5 No 9 (2021): September 2021
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

In December 2019 in the city of Wuhan, China, a case known as coronavirus disease 2019 (Covid-19) emerged and spread rapidly throughout the world. The Indonesian government implements a distance learning policy (PJJ) to minimize the spread of Covid-19. Opinions about PJJ are conveyed by the public via tweets. Emotion detection is the process of classifying tweets into emotion classes. Term weighting is a basic problem in text classification because it can affect accuracy. TF-IDF is one of the most frequently used term weightings, but TF-IDF is not the most effective because it ignores class labels. Therefore, emotion detection in tweets is carried out in order to find out emotions about PJJ. In this study, emotion detection will go through several processes, namely preprocessing, weighting of the Term Frequency-Inverse Gravity Moment (TF-IGM), cosine similarity, classification using the K-Nearest Neighbor (KNN) method, and evaluation using confusion matrix. Based on the test results using an imbalanced dataset, the optimal TF-IGM weighting coefficient is 9 which produces the highest accuracy of 0.55 at k = 25. The use of the TF-IGM weighting coefficient provides an accuracy that is less stable when compared to the TF-IGM without the weighting coefficient. The weighted words TF-IGM and TF-IDF have the same highest accuracy value, and the distance between evaluation results is small for each k tested.
Co-Authors Imam Cholissodin Yuita Arum Sari