ReCloud: Semantics-Based Word Cloud Visualization of User Reviews

User reviews, like those found on Yelp and Amazon, have become an important reference for decision making in daily life, for example, in dining, shopping and entertainment. However, large amounts of available reviews make the reading process tedious. Existing word cloud visualizations attempt to provide an overview. However their randomized layouts do not reveal content relationships to users. In this paper, we present ReCloud, a word cloud visualization of user reviews that arranges semantically related words as spatially proximal. We use a natural language processing technique called grammatical dependency parsing to create a semantic graph of review contents. Then, we apply a force-directed layout to the semantic graph, which generates a clustered layout of words by minimizing an energy model. Thus, ReCloud can provide users with more insight about the semantics and context of the review content. We also conducted an experiment to compare the efficiency of our method with two alternative review reading techniques: random layout word cloud and normal text-based reviews. The results showed that the proposed technique improves user performance and experience of understanding a large number of reviews.


Ji Wang , Department of Computer Science, Virginia Tech
Jian Zhao , Department of Computer Science, University of Toronto
Sheng Guo , Department of Computer Science, Virginia Tech
Chris North , Department of Computer Science, Virginia Tech
Naren Ramakrishnan , Department of Computer Science, Virginia Tech


ACM DL link:


GI 2014, Montreal, Canada:


Yelp Academic Dataset:


Grand Prize ($5,000) in Yelp Dataset Challenge.

InfoVis Taxomony: