The STViewer, a Visual Method with Sentiment Analysis
Wenjun Wang, Feng Zhou, Wei Xiao · 2019
In this paper, we propose a new visualization method called STViewer, which can visualize a large number of social media text based on a certain topic. STViewer can not only perceive the sentiment polarity of the whole text, show the opinions with high frequency of occurrence, but also preserve the semantic structure of the unstructured text. This helps users gain an overview of the key concepts and opinions of a large-size text rapidly and visually. Our method also provides users with some interactions to know more about his interested details. The method combines the design ideas of word cloud and word tree, which presents the popular opinions with large font and use some links to show the structures between words, and overcomes the limitations of both the two. STViewer is implemented as a light-weight application runs in the browser. Additionally, we propose a sentence sentiment analysis algorithm which is suitable for light-weight application and has good performance.