COMPANY PROFILE
Semasphere engineers components and applications to help identify meaningful and valuable information in large volumes of data through a process which extensively uses domain and expert knowledge to guide the data analysis. Our approach targets complex problem domains, where uncovering meaningful and valuable information in large data collections cannot be achieved without using the internal semantics of the domain and without human domain expertise.
The approach we are using has been successfully tested at application level in complex engineering domains with large data volumes, where the understanding of events and properties burried in data is challenging. Our aim is to expand this approach, initially devised for relatively limited and very specific knowledge extraction challenges, into a generic technology usable across a multitude of high-value domains. We also aim to develop a user interaction model which is intuitive and allows domain experts with limited data analytics training, but with a deep knowledge of their problem domain, to conduct or play a leading role in the data analysis process.
Our approach is innovative in that it applies domain knowledge to a complex, high-value task without requiring a very costly knowledge encoding process. Uninformed analysis in large and complex data sets can produce voluminous and daunting collections of patterns, relations and events. A well-informed analysis, on the other hand, can be costly in terms of providing and encoding the guiding knowledge for the search process. Our work is aimed at simultaneously meeting these two challenges for the purpose of a powerful relevant information/knowledge discovery in large data sets.
In order to shape our technology to cover an extended range of potential uses, we are seeking problem 'owners' which are interested and ready to share highly relevant problem descriptions, and who are willing to partner in the specification and testing process. We invite you to read our Technology sections to get acquainted with the types of data analysis problems we are targeting, the approach we are using, the interaction between the user and platform under development and the application types we aim to support. If you are interested in exploring collaboration opportunities do not hesitate to contact us.