Some methods for predicting recommendations for MEMS designer communities

Abstract

In this article methods and algorithms for predicting recommendations for MEMS designers communities are considered. Methods use the qualification profile of the of MEMS designers. The search for MEMS designers communities two stages. In the first stage, the MEMS designers communities are searched for qualification profile of the of MEMS designers. In the second stage, MEMS designers communities are searched for the vectors of rating profiles. A method for improving the accuracy of calculating the similarity coefficients in the method of collaborative filtering is considered.