Analysis of Convolutional Neural Network for Lifelong Learning on Indonesian Sentiment Analysis

Sentiment analysis is a process to obtain the tendency of the authors in an article. Sentiment analysis classifies textual data into a class of positive, negative, or neutral sentiments. CNN is one of the deep learning algorithms capable of classifying textual data into positive, negative, or natural classes. In general, the standard learning methods learn from one domain to produce a model. Another learning paradigm is lifelong learning which is believed to be able to accumulate learning from various domains for learning in the new domain. In this paper, we examine lifelong learning of CNN for sentiment analysis on Indonesian textual data. Our simulation shows that the accuracy of CNN increases with the increase in the number of source domains where CNN learns. This shows that lifelong learning using CNN works well for sentiment analysis on Indonesian textual data.