Method and device for identifying data

Abstract

The invention discloses a method and device for identifying data, and relates to the technical field of computers. One specific embodiment of the method comprises the steps of performing word segmentation on multiple pieces of to-be-identified data to obtain to-be-identified words to generate a word vector set and a word frequency set of the to-be-identified words; matching a word vector and a word frequency of the target to-be-identified data from the word vector set and the word frequency set of the to-be-identified words respectively; wherein the target to-be-identified data is any one of aplurality of pieces of to-be-identified data; inputting the word vector of the target to-be-recognized data into a pre-trained recognition model to obtain a theme and emotion of the target to-be-recognized data; and obtaining an emotional tendency value of the target to-be-identified data according to the word frequency of the target to-be-identified data, the positive word set and the negative word set. According to the embodiment, the accuracy of data identification is improved.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a method and device for identifying data. Background technique [0002] Existing techniques for identifying data include: unsupervised Bayesian models, short text topic models, and short text sentiment topic models. Unsupervised Bayesian models and short text topic models cannot recognize the sentiment of data, but short text sentiment topic models can recognize the sentiment of data. [0003] In the course of realizing the present invention, the inventor finds that there are at least the following problems in the prior art: [0004] The short text emotional theme model is to obtain word pairs based on training data, and then obtain a model based on word pairs, so as to use the model for recognition. The words in a word pair may not belong to the same topic, or they may be words with opposite emotional polarities. Moreover, only the emotional polarity can be identified accord...

AI Technical Summary

Problems solved by technology

The words in a word pair may not belong to the same topic, or they may be words with opposite emotional polarities, and only the emotional polarity can be identified according to the word pair, and the degree of emotion cannot be determined

Therefore, the noise of the word pairs causes the prior art to have the problem that the accuracy of data recognition is not high

Images