News Sentiment Tracker: A Targeted Opinion Mining Interface
By Andrew Wesson, Prashanth Rao
Imagine you are a marketing professional responsible for managing a company’s brand. With more than 2.5 billion people reading online or print news articles on a near-daily basis, news coverage of your brand is absolutely critical. Wouldn't it be nice if you could monitor positive or negative sentiment expressed by the press without having to read hundreds of articles? Big data master's students Andrew Wesson and Prashanth Rao developed an intelligent, automated system that analyzes news content about a target (such as a person, event, product, or organization) and provides the end-user with insights on overall news sentiment toward the target, sentiment shifts over time, similarities between different publications, daily article counts, and more. Wesson and Rao pride themselves in having developed a flexible, scalable, and meaningful data product capable of fine-grained sentiment analysis by applying novel combinations of techniques from big data, natural language processing, and time series visualization to provide users with an intuitive, reliable sentiment tracking tool. The team foresees numerous commercial applications that could help guide relevant personnel in making data-driven decisions.