pictures speak a thousand words
In August 2006 along with the war in Lebanon and the terrorist bomb at Heathrow airport, breaking news in JonBenet's case was a global media lead used in a case study for EventRiver. EventRiver is a prototype of a novel visual analytics approach to streaming text exploration. In other words, information search and delivery on steriods in a user friendly, organized and easy to understand presentation of huge chunks of news media --
EventRiver: Interactive Visual Exploration of Streaming Text
Symposium on Visualization 2009
BY Dongning Luo (1), Jing Yang(1), Jianping Fan (1)
William Ribarsky(1) and Hangzai Luo(2)
1: University of North Carolina at Charlotte, USA
and 2: East China Normal University, China
Abstract
Text that keeps arriving continuously over time, such as news corpora, e-mails, and weblogs, is called streaming text. Analytical exploration of streaming text has many important applications such as government intelligence, business intelligence, and personal information management. Existing streaming text retrieval and exploration approaches suffer from the scale and the temporal complexity of streaming text. In this paper, we offer a novel streaming text visual analysis approach named EventRiver to address the above challenges. With a novel streaming text clustering algorithm aggregating text into clusters, underlying events driving the topic changes in streaming text are automatically detected and visually presented in an intuitive event visualization. A rich set of navigation, search, and inspection tools are provided to allow users to effectively browse a stream, search events of interest, and conduct detailed visual analysis. A working prototype of EventRiver has been implemented for exploring combined closed caption and broadcast news video streams. A formal user study and case studies proved that EventRiver depicts the flow of large amount of
news over time in an intuitive and understandable way.
"...Case Study 2: Detail Inspection - Second, we decided to investigate the Jonbenet Ramsey event group in more details. We selected all events in this group using the search events by group function and opened a storyboard for the selected events. The snippets in the shoeboxes showed that the event “a suspect named John Mark Karr was arrested in Bangkok, Thailand” on Aug.15 triggered new developments in the unsolved “Jonbenet Ramsey Murder Case”: On Aug.17, Karr admitted being in the company with Jonbenet when she was killed; Karr was transferred to Los Angeles (CA) from Thailand on Aug.20; Karr waived extradition in Los Angeles County Superior Court, clearing the way for his transfer to Boulder on Aug.22; and Karr was transferred to Boulder (CO) for trial on Aug.24. (Subsequently, it was found he lied about his involvement.)..."
EventRiver: Interactive Visual Exploration of Streaming Text
Symposium on Visualization 2009
BY Dongning Luo (1), Jing Yang(1), Jianping Fan (1)
William Ribarsky(1) and Hangzai Luo(2)
1: University of North Carolina at Charlotte, USA
and 2: East China Normal University, China
Abstract
Text that keeps arriving continuously over time, such as news corpora, e-mails, and weblogs, is called streaming text. Analytical exploration of streaming text has many important applications such as government intelligence, business intelligence, and personal information management. Existing streaming text retrieval and exploration approaches suffer from the scale and the temporal complexity of streaming text. In this paper, we offer a novel streaming text visual analysis approach named EventRiver to address the above challenges. With a novel streaming text clustering algorithm aggregating text into clusters, underlying events driving the topic changes in streaming text are automatically detected and visually presented in an intuitive event visualization. A rich set of navigation, search, and inspection tools are provided to allow users to effectively browse a stream, search events of interest, and conduct detailed visual analysis. A working prototype of EventRiver has been implemented for exploring combined closed caption and broadcast news video streams. A formal user study and case studies proved that EventRiver depicts the flow of large amount of
news over time in an intuitive and understandable way."...Case Study 2: Detail Inspection - Second, we decided to investigate the Jonbenet Ramsey event group in more details. We selected all events in this group using the search events by group function and opened a storyboard for the selected events. The snippets in the shoeboxes showed that the event “a suspect named John Mark Karr was arrested in Bangkok, Thailand” on Aug.15 triggered new developments in the unsolved “Jonbenet Ramsey Murder Case”: On Aug.17, Karr admitted being in the company with Jonbenet when she was killed; Karr was transferred to Los Angeles (CA) from Thailand on Aug.20; Karr waived extradition in Los Angeles County Superior Court, clearing the way for his transfer to Boulder on Aug.22; and Karr was transferred to Boulder (CO) for trial on Aug.24. (Subsequently, it was found he lied about his involvement.)..."
Labels: EventRiver, John Mark Karr, JonBenet, JonBenet Ramsey
<< Home