RevMap: A Visualized Framework for Holistic View of Reviews
Shekoofeh Mokhtari, Tao Li, Ning Xie · 2018
Our daily digital life is surrounded by algorithmically selected contents such as recommendations, reviews, and news feeds. The process of extracting useful information out of large volumes data is imposing great impacts on many aspects of people's life. However, as the data volume becomes huge, users tend to get lost in details and miss the general picture. Indeed, the fast-growing? online data has required the development of systems that not only accurately distill useful and representative knowledge out of large data corpus but also display the extracted information in an easy-to-understand manner. In this paper, we propose a unified framework which generates visual structured summaries of customer reviews, thus providing a holistic view of a large group of reviews. Our model employs a deep neural network for opinion mining which relies on the character-level inputs. To the best of our knowledge, previously there is no uniform framework that performs visual summarization of consumer opinions as proposed in this paper.