Translating frame model construction with use of terms clustering

The knowledge representation frame model in IP-systems of machine translation is considered. The segmentation algorithm of the source and target text through communication is offered. Various terms clustering methods are analysed, the most suitable are offered to clusters text splitting.