In the information technology world, recognition systems refers to computer programs that teach a computer to perform pattern recognition, where patterns can be images, waveforms, or sets of measurements. These programs give computers a huge range of their current abilities, from speech recognition to biometric identification to prediction modeling. According to an article in Expert Systems this is important because we live in a 'knowledge economy," where the analysis, transfer, and utilization of knowledge is vital to keeping things running.
The definition of "recognition systems" overlaps heavily with that of "pattern recognition," "machine learning," and "data mining." It is difficult to define one without defining the others. However, generally, machine learning and recognition systems are used in settings which rely on artificial intelligence, whereas data mining refers to the analysis of data, usually in a business setting. Bear in mind that this is a Wikipedia definition and may not be in line with current official thinking.