Analyzing Concerns of People from Weblog Articles Tomohiro FUKUHARA 1 , Toshihiro MURAYAMA 1 and Toyoaki NISHIDA 2 1 Research Institute of Science and Technology for Society {fukuhara,tmuraya}@ristex.jst.go.jp 2 Department of Intelligence Science and Technology, Graduate School of Informatics, Kyoto University nishida@i.kyoto-u.ac.jp Abstract A system for analyzing concerns of people from Weblog articles is proposed. The system called Kanshin analyzes collective and personal concerns by collecting Weblog articles. The system collects RSS (RDF Site Summary) files of Japanese Weblog sites. The system provides keywords of the day and the month. Several patterns of collective and personal concerns are described. 1 Introduction Understanding concerns of people is important for understanding a society. In the city of Pompeii which is one of cities of the Roman empire, various graffiti are found on walls 1 . In the Roman era, there were various problems as well as today. From the graffiti, we can find personal and social concerns of people. Today, understandingconcerns of people is important for solving social problems. There are many social problems in our society. For example, we have concerns over BSE (Bovine Spongiform Encephalopathy), SARS (Severe Acute Respiratory Syndrome), GMO (Genetically Modified Organism), and so on. For tackling with these problems, understanding concerns of people is important for finding key points of the problems to be solved. We propose a system for understanding concerns of people from Weblog articles. Because Weblog has become one of important information channels to publish our thoughts and ideas on the Internet, we collect and analyze Weblog articles for finding concerns of people from collective and personal viewpoints. Users of this system can understand current concerns of people on his or her Web browser instantly. This paper consists of following sections. In Section 2, we describe the aim of this research, and requirements for the prototype system. In Section 3, we describe an overview of the prototype system called Kanshin. In Section 4, we describe several patterns of social (collective) concerns found by the prototype system. In Section 5, we describe an approach to find calm words which have been topic-indicating words but mentioned rarely in recent articles, and some examples of calm words. In Section 6, we describe an approach and analysis results of personal concerns. In Section 7, we discuss differences between our system and other works. In Section 8, we summarize arguments of this paper, and describe the future work. 2 Analyzing concerns of people from Weblog articles In this section, we describe (1) the aim of this research, and (2) requirements for the prototype system. 1 http://www.noctes-gallicanae.org/Pompeii/graffiti 1.htm (in French; accessed February 15, 2005)