An Information Visualization System to Assist News Topics Exploration with Social Media

With the popularity of social media, journalists often collect news materials from mass user-generated contents. However, with the increase of social media data, it is not easy for a journalist to see the whole picture of an event from the huge amount of data. If they only use the social media as a news source, the reported content may often become a copied view of the users, or fall into the stereotype of one-sided discussions. Aiming to find a solution to this problem, our study uses Twitter data as an example to develop an information system to assist journalists to explore events, collect materials, and find news topics through social media. We use network analysis and natural language processing techniques to analyze the collected data and visualize the story elements. We have developed four story elements models to support different ways of exploring the data. We let the users adjust the weight on the importance of these models to show the context of tweets and help users find news topics. We have designed a two-phase experiment with questionnaires to assess the appropriateness of the system. We allow the participants with various degrees of familiarity with the event to explore news topics on our system. The experimental results show that the participants have found the system to be useful and easy to use, and the journalists can explore news topics and track events in a much faster fashion.