Predict future outcomes by having UBIC’s
Virtual Data Scientist analyze historical events/facts.
In the world of data analysis, an engineer that can read some trends from changes in innumerable data, a so-called “data scientist,” is necessary, but there are concerns about the lack of these human resources. The behavior informatics of UBIC fulfills the role of a true “Virtual Data Scientist” by analyzing and predicting the thoughts of people.
In eDiscovery (electronic evidence disclosure), the evidence disclosure system in United States litigation, it is necessary to find from among vast amounts of electronic data the electronic documents that constitute evidence pertaining to the case and disclose them in a predetermined period of time. The amount of electronic data is increasing year by year and investigations relying on the method of employing large numbers of human investigators such as lawyers, etc. to process that data have approached their limit. Therefore, we concluded that a software investigator to replace humans (Predictive Coding) was necessary.
In addition, the electronic documents that the investigators want to find vary depending on the type of case (patent infringement litigation, anti-monopoly laws, etc.) and the circumstances unique to companies. In order to deal with these issues UBIC has succeeded in developing Predictive Coding, which is more flexible, offers more efficient calculations, and is suitable for making future predictions.
UBIC has been developing a “Virtual Data Scientist,” a software expert that can further accumulate and utilize knowledge. The virtual data scientist is capable of performing not only conventional clustering and trend analyses of vast amounts of electronic data but also analyses that utilize findings with respect to humans and society to a greater extent, which is a qualitative approach based on databases of case studies and the products of behavior science, and we will widely utilize it for visualizing and predicting the behavior of humans and communities, areas which reside in big data.