A news comment automatic generation method and device

Abstract

The invention relates to a news comment automatic generation method and device. The method comprises the steps of taking a news title as original data; preprocessing the news title data to construct training data, wherein the training data is used as input of an LSTM model to train the model; obtaining a pre-trained LSTM model through training the sample data; adding a discriminator into the model; forming a GAN model, and distinguishing the comments generated by the LSTM model from true comments and false comments by using a discriminator, using the LSTM model as a generator to be in game with the discriminator until an output result reaches a threshold value. In order to ensure the context correlation between the output result (comments generated by the model) and news, a gating attention mechanism is applied to effectively process the context information.

Description

technical field [0001] The invention relates to the technical field of computer networks, in particular to a method and device for automatically generating news comments. Background technique [0002] The classic network is the recurrent neural network (RNN), which is also the network of choice for time series data. When it comes to certain sequential machine learning tasks, RNNs can achieve high levels of accuracy that no other algorithm can match. This is due to the fact that traditional neural networks only have a kind of short-term memory, while RNN has the advantage of limited short-term memory. [0003] The purpose of RNNs is to process sequence data. In the traditional neural network model, from the input layer to the hidden layer to the output layer, the layers are fully connected, and the nodes between each layer are unconnected. But this ordinary neural network is powerless for many problems. For example, if you want to predict what the next word in a sentence ...

AI Technical Summary

Problems solved by technology

Previous deep generative models required Markov chains or approximate maximum likelihood estimation, resulting in many difficult-to-calculate probability problems. They could not score partial sequences, but could only evaluate complete sequences. For discrete data, backpropagation has problems.

Images