Data flow processing framework for multimodal data environment software

Abstract

Presented is a general purpose, very efficient data processing platform Multimodal Data Environment (MMDE), that includes a new framework for flexible data analysis and processing - Data Flow Programming Framework (DFPF). It was designed to unify already developed solutions, integrate them together and reuse as much of them in the simplest possible way. At the lowest level DFPF provides wrappers for current code to run it within a new framework and at the higher level it is possible to visually construct new algorithms for data processing from previously created or provided processing elements according to visual programming paradigm. DFPF is implemented as a dedicated service for MMDE in C++, but we expect that such solution can be easily implemented in any modern programming language.