The design of electronic medical records system using Skip-gram algorithm

Abstract

To improve the management ability of patient information and establish a feasible electronic medical record (EMR) management system, combined with the characteristics of e-commerce, the EMR system is established. Moreover, based on Skip-gram algorithm, the word database of EMR is segmented and extracted. First, the data in the EMR are preprocessed to extract the specific information from the EMR. Then, the data preprocessed based on Skip-gram algorithm are analyzed and processed to realize the automatic naming, recognition and annotation of EMR. The experimental results show that the precision, recall and F1-measure of Skip-gram model are 87.83%, 87.25% and 87.54%, respectively, which are better than those of RNN model. Therefore, Skip-gram algorithm is feasible and effective in the construction of EMR system. This exploration provides a reference for data processing of EMR system.