Automatic text classification method for power grid user fault repair work orders

Abstract

The invention relates to an automatic text classification method for power grid user fault repair work orders. The automatic text classification method comprises the steps: firstly, conducting word segmentation on data of a text data set of a power grid user fault repair work order, filtering stop words and training word vectors, wherein a Word2vec method is adopted to train word vectors and Word2vec learns distributed expression of the text, vectorizing the text, and expressing the text through the word vectors; and loading the trained word vectors into a convolutional neural network for feature extraction and model training, and finally outputting probability distribution on each label to obtain a classification result. Compared with a traditional text classification method, the automatic text classification method obviously improves the accuracy, and greatly shortens the classification time. Meanwhile, the automatic text classification method can help the power grid department morereasonably configure various emergency maintenance resources, thus greatly shortening the fault maintenance time, and facilitating the improvement of the service quality of the power grid department.

Description

technical field [0001] The invention relates to a text classification technology, in particular to a text automatic classification method for a power grid user's fault repair work order. Background technique [0002] In recent years, with the rapid development of economy and informatization, the scale of the power grid has been continuously expanding, and the number of users has also been increasing. Power grid user fault repair work orders are important information for recording user feedback. These information are mainly presented in the form of Chinese text, with short content and sparse features. Traditionally, power grid fault repair work orders mainly rely on manual classification. However, this processing method is not only inefficient, but also extremely error-prone. Contents of the invention [0003] The invention aims at the problem that it is difficult for the current power grid department to automatically process and classify user fault repair work orders, an...

AI Technical Summary

Problems solved by technology

[0003] The present invention is aimed at the problem that it is difficult for the power grid department to automatically process and classify user fault repair work orders, and proposes a text automatic classification method for power grid user fault repair work orders to realize automatic classification

Images