Method and system for automatic recognition of medical text terms based on long-term and short-term memory network

Abstract

The invention discloses a method and a system for automatic recognition of medical text terms based on long-term and short-term memory network, which are designed for automatically extracting medicalterm entities from medical text. The method comprises the following steps: each character in a medical text sentence is represented by a pre-trained character vector to obtain training data; the training data is inputted into the two-way long-short memory network to obtain the label category with the highest probability of each character in the medical text sentence; the output is input into a conditional random field and the Viterbi algorithm is used to calculate the label sequence with the highest probability of association. The method integrates the respective advantages of the two-way long-short-time memory network and the conditional random field, and can effectively improve the accuracy rate of word labeling.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to view more

AI Technical Summary

Benefits of technology

This technology uses both 2G/3G or Long Term Memory Network (LTM) techniques for improving speech processing systems by combining different aspects from previous methods such as LLMARA models into one framework. These improvements help identify specific parts of an audio signal better than other approaches like AMR modeling.

Problems solved by technology

This patented technical problem addressed by this patents relates to improving automated medicine recognizing technology with regards to both word matchings (TNC) or autoimmune terminology identification techniques such as neural network models.

Images