Emotion discrimination method based on fine-grained annotation data

A technology for labeling data and discriminating methods, applied in the field of text processing, can solve problems such as increased labeling workload, and achieve the effect of less classification data and better effect

CN111046171AActive Publication Date: 2020-04-21CHENGDU UNIV OF INFORMATION TECH
7 Cites 3 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Publication Date
2020-04-21

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
Patent Text Reader

Abstract

The invention relates to an emotion discrimination method based on fine-grained annotation data. The method comprises the following steps: collecting financial news data, dividing news data into a labeled sample set and an unlabeled sample set; training a first classifier and a second classifier through the labeled sample set and the unlabeled sample set; the first classifier can screen out the key sentences in the article; the second classifier judges the emotional tendency of the article; model parameters of the first classifier and model parameters of the second classifier are obtained respectively; and adding the data with high confidence in the classification result into the annotation sample set, selecting the most worthy annotated data C from the unannotated sample set by using an active learning theory, sending the most worthy annotated data C to a worker for annotation, and circularly training the emotion discrimination model until the classification precision is achieved, andending the training to obtain the discrimination model.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to the field of text processing, in particular to an emotion discrimination method based on fine-grained labeled data. Background technique

[0002] In the current era of information overload, the speed of news generation far exceeds the speed that individuals can process. In order to ensure that users can obtain effective information, proper feature extraction and filtering of original news has become a relatively common and necessary practice. When using mathematical models to quantify financial news, the emotional tendency (positive / negative / neutral) of the news is one of the very important attributes.

[0003] There are three ideas for sentiment classification of news texts: sentiment lexicon-based methods, machine learning-based methods, and deep learning-based methods.

[0004] The method based on sentiment lexicon mainly judges the sentiment polarity of the text by constructing sentiment lexicon and a series of rules; from...

Examples

Embodiment Construction

[0026] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in combination with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are exemplary only, and are not intended to limit the scope of the present invention. Also, in the following description, descriptions of well-known structures and techniques are omitted to avoid unnecessarily obscuring the concept of the present invention.

[0027] The fine-grained labeling data in the present invention refers to: when carrying out emotion classification to chapter in traditional method, all is to only label the emotion polarity label of whole article only, when training emotion classification model, all is to use whole article as The input of the model, the emotional polarity label of the article is used as the output of the model, and then the training and ...