Fujio Toriumi

Information sharing on Twitter during the 2011 catastrophic earthquake

Such large disasters as earthquakes and hurricanes are very unpredictable. During a disaster, we must collect information to save lives. However, in time disaster, it is difficult to collect information which is useful for ourselves from such traditional mass media as TV and newspapers that contain information for the general public. Social media attract attention for sharing information, especially Twitter, which is a hugely popular social medium that is now being used during disasters. In this paper, we focus on the information sharing behaviors on Twitter during disasters. We collected data before and during the Great East Japan Earthquake and arrived at the following conclusions: Many users with little experience with such specific functions as reply and retweet did not continuously use them after the disaster. Retweets were well used to share information on Twitter. Retweets were used not only for sharing the information provided by general users but used for relaying the information from the mass media. We conclude that social media users changed their behavior to widely diffuse important information and decreased non-emergency tweets to avoid interrupting critical information.

Tweet trend analysis in an emergency situation

The Great Eastern Japan Earthquake, which struck Japan on March 11, catastrophically affected all aspects of life: buildings, power plants, human life, etc. Moreover, it caused severe problems related to network infrastructure. We can ascertain the degree of network disorder from network traffic logs. Although we can infer what people did when the earthquake occurred on the Web from network traffic logs, we cannot know it precisely. Social media were used effectively during and after this earthquake, and they left a partial log revealing what people did on the Web during and after the earthquake. Such a log is one of the first logs of peoples actions in a time of a catastrophic disaster. As described in this paper, we analyze Twitter logs and attempt to extract what happened in the emergency situation.

Evaluation of automated-trading strategies using an artificial market

The purpose of this study was to extend the simulation of artificial markets to the practical application. We constructed an artificial-market system with an interface with the automated-trading strategies. Then, using an artificial-market simulation, we conducted two types of evaluations of automated-trading strategies that could participate in a Kaburobo competition. (1) Our system evaluated the risks and returns of the strategies in various market environments. An evaluation using the artificial market was able to provide better information than a conventional evaluation using a back test. (2) Our system could also test the market impact of automated-trading strategies. Our results revealed that the market impact of the strategies may not only depend on their rule content but also on the way they are combined with other strategies.

Information Diffusion on Twitter: Everyone Has Its Chance, But All Chances Are Not Equal

Twitter is a Web 2.0 social network which attracted much attention recently for its usage as an alternative media for information diffusion. From the recent events in Arab countries, to natural disaster such as earthquakes or tsunamis, Twitter has proven to be a credible alternative to traditional means of information diffusion. Relatively few works have been done on this question of information diffusion, and in particular on the relative importance of different kind of users on this question. In this paper, we show that all users are not equal on the aspect of information diffusion. By investigating thoroughly the retweet chain lengths of users on a large dataset, we found that the number of followers of users plays an important role in their capacity to propagate information. From our observations we propose a very simple model, which is accurate enough to generate realistic length of retweet chains on the network. We consequently show, by studying a Twitter dataset centered on the Japanese Earthquake and Tsunami in March 2011, that such a crisis impact greatly the propagation of information. Finally, we use our results to discuss on the means of improving information diffusion to reach targeted users.

The possibility of social media analysis for disaster management

Collecting, sharing, and delivering information in disaster situations is crucially important. Mass media such as TV, radio, and newspapers have played important roles in information distribution in past disasters and crises. Recently, social media have received much attention for their use as an information sharing tool. Especially, it is said that people used Twitter to collect and share information in the aftermath of the Great East Japan Earthquake. In academic fields, some researchers have started to propose some methods and systems for disaster management by analyzing social media data. Other people doubt whether social media will actually function effectively for disaster management because of uncertainty and inaccuracies related to rumors and misunderstanding. In this paper, we overview current studies of social media analysis for disaster management and explain some studies in detail to show their possibility and availability. We specifically examine situational awareness, user behavior analysis and information propagation analysis, which are three approaches to social media analysis, to clarify what social media analysis can and cannot do. Additionally, we propose some concepts for social media analysis and show how those concepts help to collaborate with us, researchers in social media analysis fields and other research fields.

Classification Method for Shared Information on Twitter Without Text Data

During a disaster, appropriate information must be collected. For example, victims and survivors require information about shelter locations and dangerous points or advice about protecting themselves. Rescuers need information about the details of volunteer activities and supplies, especially potential shortages. However, collecting such localized information is difficult from such mass media as TV and newspapers because they generally focus on information aimed at the general public. On the other hand, social media can attract more attention than mass media under these circumstances since they can provide such localized information. In this paper, we focus on Twitter, one of the most influential social media, as a source of local information. By assuming that users who retweet the same tweet are interested in the same topic, we can classify tweets that are required by users with similar interests based on retweets. Thus, we propose a novel tweet classification method that focuses on retweets without text mining. We linked tweets based on retweets to make a retweet network that connects similar tweets and extracted clusters that contain similar tweets from the constructed network by our clustering method. We also subjectively verified the validity of our proposed classification method. Our experiment verified that the ratio of the clusters whose tweets are mutually similar in the cluster to all clusters is very high and the similarities in each cluster are obvious. Finally, we calculated the linguistic similarities of the results to clarify our proposed methods features. Our method classified topic-similar tweets, even if they are not linguistically similar.

Evolution of cooperation in SNS-norms game on complex networks and real social networks

Social networking services (SNSs) such as Facebook and Google+ are indispensable social media for a variety of social communications, but we do not yet fully understand whether these currently popular social media will remain in the future. A number of studies have attempted to understand the mechanisms that keep social media thriving by using ameta-rewards game that is the dual formof a public goods game. However, the meta-rewards game does not take into account the unique characteristics of current SNSs. Hence, in this work we propose an SNS-norms game that is an extension of Axelrod’s metanorms game, similar to meta-rewards games, but that considers the cost of commenting on an article and who is most likely to respond to it. We then experimentally investigated the conditions for a cooperation-dominant situation in which many users continuing to post articles. Our results indicate that relatively large rewards compared to the cost of posting articles and comments are required, but optional responses with lower cost, such as “Like!” buttons, play an important role in cooperation dominance. This phenomenon is of interest because it is quite different from those shown in previous studies using meta-rewards games.

Investigation of the effects of nonverbal information on werewolf

Werewolf is one of the popular communication games all over the world. It treats ambiguity of human discussion including the utterances, gestures and facial expressions in a broad sense. In this research, we pay attention to this imperfect information game werewolf. The purpose of the research is to develop an intelligent agent “AI werewolf” which is enabled to naturally play werewolf with human. This paper aims to investigate how behavior contribute to victory of own-side players by using machine learning as a first step. As the results of investigation and analysis of the playing movie, we found that nonverbal information in the game of werewolf has importance to winning or losing the game.

Evolution of Cooperation in Meta-Rewards Games on Networks of WS and BA Models

We investigated the required conditions in which cooperation is dominant in social media using the model of a meta-rewards game, which is a dual part of Axel rods metanorms game. Social media such as Twitter and Facebook have rapidly been growing in recent years. However, we do not know whether or not the currently popular social media will remain in the future. A number of studies have been conducted to try to understand the conditions or mechanisms that create and keep social media thriving using a public goods game and/or meta-rewards games, in which situations where many users post articles and respond to them as reactions in social media correspond to situations where cooperation is dominant in these games. However, they assume that agent networks in social media are complete graphs that are known to be dissimilar to actual social networks. We examined the conditions required to keep agent networks thriving based on Watts and Strogatz (WS) and Barabasi-Albert (BA) models, which are more similar to actual social networks. We experimentally found similarities and differences in the conditions for cooperation-dominant situations between networks based on complete graphs and those based on WS and BA models. Our results indicated that it was easier to maintain cooperation-dominant situations in BA-model networks than in other networks.

Great east Japan earthquake viewed from a URL shortener

On March 11th 2011, a great earthquake and tsunami hit eastern Japan. After that, several web sites, especially those providing helpful disaster-related information, were overloaded due to flash crowds caused by Twitter users. In order to mitigate the flash crowds, we develop a new URL shortener that redirects Twitter users to a CDN instead of original sites, since Twitter users rely on URL shorteners like bit.ly to shorten long URLs. In this paper, we describe our experience of developing and operating the URL shortener in the aftermath of the giant earthquake. Since the flash crowds were a serious problem in an emergency, we had to develop it as quickly as possible with a spirit of so-called agile software development. We then explain our HTTP request log collected at the URL shortener (it is now available online). To investigate the cause of flash crowds, the log is examined with tweets (Twitter messages) provided by another research project; this collaboration is realized by the encouragement of the workshop committee. We hope our experience will be helpful in tackling future disasters.