Proposal of Social-Mass Media Triplification and Its Use Case

Takahiro KAWAMURA  Kenji KOSHIKAWA  Hiroyuki NAKAGAWA  Yuichi SEI  Yasuyuki TAHARA  Akihiko OHSUGA  

D - Abstracts of IEICE TRANSACTIONS on Information and Systems (Japanese Edition)   Vol.J96-D   No.12   pp.2987-2999
Publication Date: 2013/12/01
Online ISSN: 1881-0225
Print ISSN: 1880-4535
Type of Manuscript: Special Section PAPER (Special Section on Software Agent and Its Applications)
agent,  linked data,  social-vs-mass media,  

Full Text(in Japanese): PDF(3.5MB)
>>Buy this Article

Growth of Internet makes easy access to several information sources such as social and mass media, and then truly diverse attitudes and opinions these day. At the same time, users are required to judge the information credibility due to spreading of false rumor in the social media and suspicion of biased coverage in the mass media. Thus, in order to support the comparison of media information from several aspects, we are aiming at an agent system which extracts and visualizes the information of a certain event from both medias, and then presents some points to be compared. This paper proposes a method to extract the event information with 13 attributes from tweets and news articles using Conditional Random Fields and heuristic rules, and convert it to Linked Data, and then evaluates its extraction accuracy. Moreover, it confirmed the usefulness though case studies by extracting comparable points from four aspects of diversity, infrequency, uneven distribution, causal association.