METHOD OF RECOGNIZING NAMED ENTITIES IN NETWORK TEXT BASED ON ELIMINATION OF PROBABILITY AMBIGUITY IN NEURAL NETWORK Russian patent published in 2020 - IPC G06F40/295 G06N3/08 

FIELD: computer equipment.

SUBSTANCE: invention relates to computer engineering. Disclosed is a method of recognizing named entities of network text based on eliminating ambiguity of probability in a neural network, involving: performing word decomposition on unmapped text body using Word2Vec model to extract word vector, converting reference text bodies into a word feature matrix, performing window processing, constructing a deep neural network for training, adding a Softmax function to the output layer of the neural network and performing normalization to obtain a matrix of probabilities of the category of named entities corresponding to each word; performing repeated processing of the probability matrix by the window method and using the model of conditional random fields to eliminate ambiguity to obtain the final tag of the named entity.

EFFECT: enabling recognition of named entities of network text based on elimination of probability ambiguity in a neural network.

7 cl, 3 dwg