Infectious disease surveillance is a public health practice to monitor the prevalence of contagious diseases in a target region. However, the conventional surveillance has recently failed at detecting diseases such as Dengue fever and Severe fever with thrombocytopenia syndrome (SFTS) that have not been recognized in Japan. The current surveillance methods mostly depend on physician reports, and thus, it is hardly possible to detect diseases that are unknown, or uncommon in the regions. Accordingly, it is highly beneficial to have a disease surveillance method that detects even unknown disorders, as well as the common ones. In this regard, the big-data approach could have been an alternative to monitor the trend of infectious dis- eases, based on behavioral information of people. Nevertheless, data source of the proposed approaches are mostly unreliable, and vulnerable to deceptions. We propose to maintain a diagnostic decision support system as a nationwide public service to physicians, in the aim of collecting search queries for hard-to-diagnose cases across the nation. The collected information would be a desirable source for disease surveillance, and anomaly detection algorithms can efficiently monitor unknown diseases. This paper reviews diagnostic systems for such purpose, and discusses the emerging application of artificial intelligence in society.

In order to solve a social problem, e.g. urban traffic jam, it is expected to analyze social phenomena using numerical simulation. This study is an early step to solve such problems. Our target phenomenon is urban vehicle traffic. An analysis is performed to make ensure that there are some relationships between traffic demand in each road. From the result of the analysis, we can show that such relationship may be existed. At the last session, we have a discussion to guarantee the possible existence of such relationship in the real world.