Detecting Emotion on Indonesian Online Chat Text Using Text Sequential Labeling
Ramos Janoah Hasudungan, Masayu Leylia Kodhra · 2018
Emotion Classification on online chat is a text classification to determine which emotion is in the text. In this paper, we employ sequential labeling with various feature, i.e. bag of words 1-gram and 2-gram, pragmatic feature, non-textual feature, and word embedding feature. We also compare sequential labelling model with non-sequential labelling model. Long Short-Term Memory (LSTM) is employed as sequential labeling technique, and several machine learning for non-sequential approaches. The best result F1-score in validation set and testing set are 0.325 and 0.326 respectively, achieved by using non-sequential approaches, with Multilayer Perceptron and 1-gram, pragmatic feature, non-textual feature and average word vector of word embedding as its feature.