Secondary equipment fault short text data classification method based on convolutional neural network

Abstract

The invention discloses a secondary equipment fault short text data classification method based on a convolutional neural network, and the method comprises the steps: firstly collecting text information data of a secondary equipment fault, determining a data set, and obtaining a training sample set; preprocessing the secondary equipment fault short text data, and performing one-to-one mapping on the word data and the word vectors to obtain text vector data; training the convolutional neural network model by using the training sample to obtain a trained convolutional neural network model; and according to the verification set data, inputting test input data into the trained convolutional neural network model, and taking a model output value as a classification result of the to-be-classifiedsecondary equipment fault short text information. According to the secondary equipment fault short text data classification method based on the convolutional neural network, the prediction precisionis improved and the generalization ability of the model is enhanced by using the good feature screening analysis ability of the convolutional neural network model.

Description

technical field [0001] The invention belongs to Chinese natural language processing technology, and in particular relates to a method for classifying short text data of secondary equipment faults based on a convolutional neural network. Background technique [0002] During the construction and operation of the smart grid, electric power big data has exploded. According to the white paper on the development of China's electric power big data compiled by the China Electrical Engineering Society Informatization Committee in 2013, these data can be roughly divided into two categories. One is based on Output power, equipment and its environment temperature and humidity, light intensity of optical modules, etc. represent time-series structured data, and the other type is semi-structured and unstructured data represented by text, images, audio, etc., which are difficult to express using relational databases. structured data. The structured data mining work has been relatively matu...

AI Technical Summary

Problems solved by technology

[0003] During the operation of secondary equipment, a lot of short-text data of faults and defects has been accumulated. These data are often recorded manually by inspection personnel to complete the classification of defects. However, due to the difference in subjectivity and experience of inspection personnel, it is difficult to Accurate classification is achieved, and due to the large amount of fault data, a large amount of manpower is required, and the efficiency is difficult to guarantee

[0004] At present, for the classification of short texts, there have been researches abroad on emotional classification of hotel reviews through natural language processing. Text classification does not work, English natural language processing is difficult to use in Chinese, and due to different industries, there are many proper nouns in the field of secondary equipment text classification, lack of research on the improvement of classification models, mostly based on traditional machine learning fields

In addition, due to manual recording by recorders, there are many colloquial records and short text information. At present, there is no public method for classifying short text information of secondary equipment.

Images