SenseTag: A tagging tool for constructing Thai sentiment lexicon

Abstract

In research of sentiment analysis, polarity terms are important elements to help classify the polarity of text to be positive or negative. This paper presents SenseTag which is a sentiment tagging tool for constructing Thai sentiment lexicon on web-based platform. It consists of three main components: data collection, data annotation and administrative operation. The first component is to collect data from social media such as Twitter and www.pantip.com, then preprocess them. The second component is designed to make it easier for linguist to annotate data. The third component allows administrator to configure and manage tag category, domain corpus and user account. We discuss and report the result of annotation in four topics (e.g., mobile phone, automobile, stock market, and general).