2011年3月11日に発生した東日本大震災の際に，Twitterが情報を伝達・収集するツールとして機能したことは記憶に新しい．そのような中，災害時の新たな情報源としてのソーシャルメディアが期待を集めている．現在，ソーシャルメディア上での情報拡散の研究が多く行われている．本研究では，情報拡散が行われるネットワークの構造に着目し分析を行う．震災によるコミュニケーションネットワークの構造変化は情報拡散のしやすさに影響を及ぼすと考えられる．そこで，震災によって変化する前後のネットワークの情報拡散能力を評価し，震災後のネットワークの方が情報拡散能力が高いことを示唆した．また，それがどのような要因によって引き起こされたものかを分析し，次数分布が重要な要因であることを示唆した．

During an emergency, information on the emergency is important to egress safely from buildings and to conduct rescue operations quickly. Nowadays, almost everyone has mobile phones and they communicate each other using social network (SNS). In this paper, we propose three types of communication model: broadcast, face to face (word of mouth), and SNS and demonstrate differences of evacuation effects.

There is social needs to know the activity type of tourists who visit to the large tourism destinations such as Kyoto. For destination management organizations (DMO), tourists classification of activity type is useful for optimal arrangement of visitor information boards, development of scenic routes, evacuation guidance and congestion mitigation. Also, DMOs can compare own itself to othter destinations by using the activity types. However, tourists classification based on the tourists activities is not studied enoughly. Therefore, we should investigate the factor of tourists activities to make a hypothesis of type classification. In this paper, we analyse the tourists' activities in large tourism destination area by transition network generated from the tourists GPS trajectories.

地域の地理的特性，人の防災に関する意識，災害経験の有無など，各地域での防災意識を高める上でのバリアや特徴とする領域は異なる．そのため，各地域での被災経験や訓練等のプラクティスが他の地域でうまく活かされないという問題が生じている．また，一般に被災地では防災意識が高い一方，近年被災していない地域では防災意識を向上させることが大きな課題となっている．この問題を解決するためには， 多地域間の情報や知識を蓄積・統合・共有する情報処理システム， 地域の事情や特性に合わせて適切に情報や知識をカスタマイズする仕組み， 効果的で科学的に検証可能な教育プログラムが必要である．本稿では，これらの問題を解決する仕組み作りの研究構想について述べ，現在の取り組みについて紹介する．

In order to improve the ability of decision making under uncertain situations during a large-scale disaster, more efficient and effective disaster response exercises are necessary. For this purpose, we developed a method to make earthquake damage scenarios semi-automatically by computer so that exercise designers can make effective exercise scenarios with a great variety without taking much time and effort. In this research, we, taking disaster nursing as an example domain, firstly made "earthquake damage model" by extracting earthquake-related elements from the reports and materials of past disasters and exercises regarding disaster nursing. Then we made "earthquake damage instance" in consideration of consistency between these elements, and finally created "earthquake damage scenario" as a text. We asked for several nurses to evaluate the prototype and obtained positive feedbacks as well as future directions to a practical application.

Many researchers launch the research in preparation using a network and information technology for the following large-scale disaster after the Great East Japan Earthquake. However, a network cannot be used immediately after at the time of a disaster. When a disaster breaks out in a strange place, it cause someone to panic and people cannot immediately cope with the situation. Furthermore, it is difficult to use an evacuation support system suddenly at the time of a disaster. Therefore, we have developed an evacuation support system, named AkariMap. AkariMap is a system for using daily, and can experience the function at the time of a disaster. We developed two new functions: the notice function which notifies a user of the refuge support information of the user's present location, and the disaster mode which can experience at the time of a disaster, From the result of the experiment, we found the followings. (1) The notice function serves as the trigger for a user to be conscious of refuge support information. (2) The disaster mode can present information required for refuge support.