Short text classification method and device

Abstract

The invention discloses a short text classification method and device, belonging to the technical field of data processing. The method comprises the following steps: obtaining short text to be classified; According to the short text traversing Huffman tree, determining the target main classification matching the short text. Wherein The Huffman tree consists of m layers of nodes, with each corresponding to a primary classification of the standard classification sample. According to the standard classification samples, the n subordinate classifications corresponding to the target master classification are determined. Acquiring short text samples corresponding to each slave classification; Calculating similarity between short text and each short text sample; According to the calculated similarities, the target classification of short text is determined. The present application solves the problems of long training period, great difficulty in implementation and more computational resourcesoccupied when the classification algorithm based on the feature analysis classifies short texts, and achieves the effects of removing the training process, reducing the difficulty in implementation and reducing the amount of computation.

Description

technical field [0001] The present application relates to the technical field of data processing, in particular to a short text classification method and device. Background technique [0002] A short text usually refers to a text consisting of 1 to 3 words, such as the name of a hospital department, the department name of a government agency, the department name of a school, etc. [0003] For classification of short texts, classification algorithms based on feature analysis, such as artificial neural networks, are mainly used in related technologies. The artificial neural network needs to be trained before it is used. The training process is as follows: Obtain a large number of training samples of known categories, input each training sample into the artificial neural network, and the artificial neural network extracts features from the training samples. According to the extracted features Predict the probability of being classified into the category of the training sample,...

AI Technical Summary

Problems solved by technology

[0005]In order to solve the problems of long training period, difficult implementation and high computing resource usage of classification algorithms based on feature analysis such as artificial neural networks to classify short texts, The embodiment of the present application provides a short text classification method and device

Images