Fujimoto Yuki

In road management, it is necessary to detect diverse phenomena such as traffic accidents and congestion and take measures including traffic regulations flexibly. Road administrators have been monitoring road traffic conditions daily, for example, through cameras installed on the roads or by patrols. However, there is a possibility that they may not be able to grasp the conditions during large-scale disasters as such equipment may break down. As a measure to deal with the situation, there are cases in recent years that road administrators analyze real-time information related to traffic phenomena posted to SNS (social network services). The contents posted to SNS contain microbloggers' purposes of travel or thoughts and impressions that are hard to collect with the present monitoring methods or statistical survey. Based on this, if a measure to utilize highly reliable post contents is established, advancement of road management can be expected. In this study, a method for analyzing reliability of microbloggers by classifying them into four quadrants is devised using the number of SNS entries posted by road users regarding a specific road and habitual behavior that can be analyzed from their past posting tendencies. The result shows that the microbloggers who are judged as reliable by the proposed method have a tendency to post useful information related to traffic phenomena. Then potential for making good use of SNS in road management is discussed using the outcome of the proposed method.

 In road traffic analysis, it has become possible to make a qualitative analysis by extracting road users' comments from microblogs, in addition to a quantitative analysis using the data obtained from diverse instruments installed on roads. Bearing utilization in abnormal situations in mind, it is necessary for the road administrator to be able to extract only highly authentic posts out of various posts instantly; however, since anyone can post microblogs freely, the authenticity of each post is unclear. This characteristic is harmful in utilizing microblogs in abnormal situations. In this research, a method for extracting the posters of authentic articles related to traffic phenomena was devised in terms of the poster's attribute and posting tendency, and its usefulness was verified.

&nbsp;Traffic phenomena including accidents occur daily on the Japanese road network. Surveillance cameras and traffic counters are used to monitor them; however, since it is fixed-point monitoring, there is a limit to the extent of monitoring. Thus, a method has been proposed for analyzing microblogs, a kind of SNS (social networking service), to complement acquisition of information concerning traffic phenomena. However, as preset keywords are used to make searches, the number of acquirable tweets fluctuates.<br>&nbsp;Therefore, this research proposes a method for selecting useful keywords for acquiring information on traffic phenomena using the characteristics of the concerned keywords to burst when a traffic phenomenon occurs, and evaluates its validity.

&nbsp;In evaluating public works projects, it is necessary to evaluate their current condition quantitatively and qualitatively from diversified viewpoints based on the social situation. For the road projects, road traffic analysis is carried out to grasp the actual traffic condition quantitatively. If this analysis result reflects public opinions that are qualitative in detail, it can be expected that this will help to advance the evaluations of road projects. To collect qualitative public opinions, one of the effective measures is to utilize tweets from Twitter, which is a form of microblogs. However, since there are all sorts of tweets in circulation, it is difficult to appropriately extract the tweets that are useful in evaluating road projects. Moreover, even if a useful tweet is extracted, it is very hard to identify the location indicated by the tweet.Thus in this research, we devised a text mining method that combines road traffic data found on the Web with tweets from Twitter, and evaluated the feasibility of extracting public opinions related to road works projects and of estimating their locations.

&nbsp;A large number of posts with useful information to help local residents can be found in CGM (Consumer Generated Media). Such posted messages contain, for example, information about prevention of crimes or emergency responses during disasters for the local communities. Various information about diverse regions is put together and open to the public. However, since anyone can contribute to CGM, it often happens that a mixture of information about multiple regions can be found on the same Web page or that different topics are covered for the same region, making it hard to extract only the information about the particular region or it's contents. This research proposes a method for extracting regional information from CGM by dividing the information circulating in CGM into posted messages, and then grouping them according to the region or it's contents that each posted message treats. It also examines a possibility of creating a regional information map by analyzing the positional information contained in the posted messages classified into groups and mapping them onto GIS.