According to the University of Tsukuba, the infrastructure system for processing stream data is for processing various data in an integrated manner and realizing various information uses, and by using this system, the cost of application development can be significantly reduced. ..In addition to being able to directly handle JSON-formatted stream data, which is becoming more widespread internationally, it is also characterized by a significant improvement in processing power when monitoring data in detail.

　We have also developed a method for estimating attributes such as location information and age of social media users with the world's highest level of accuracy.Location information and attributes are important supplementary information when interpreting the contents, but they are not explicitly disclosed.Therefore, while establishing a method for estimating with high accuracy based on the utterance content, we are also developing a method for estimating unknown user attributes based on the relationships between users and known user attributes.

　In addition, we have succeeded in developing a system that realistically matches online news and Twitter with high accuracy.By linking news articles with Twitter reactions related to them in real time, it will be possible to easily summarize the high level of public interest in news and various opinions.

　Research and development of data integration and analysis technology for the utilization of big data in the real world is a project that the Ministry of Education has been promoting for four years from 2014, with Tsukuba University as the representative school, the University of Tokyo, Tohoku University, and Keio University. A total of four schools are collaborating to continue technological development.

(* 1) Stream data is regarded as "flowing", data that flows in is input, and data that flows out is treated as output.There are various types such as those that handle file input / output and those that handle network communication.