An Emotional Classification Study Of Cambodian News Commentary Texts

This paper focuses on Cambodian journalism and the study of emotional classification.This article examines the ultimate goal of the emotional tendencies of the Cambodian news chapter level.First,we studied the level of sentence emotion classification,and produced three specific sentence emotion classification models.On the basis of the study of emotional classification,at the sentence level,through the sentence emotion classification,the emotion classification algorithm is studied at the document level from various aspects.Through the research and improvement of several main algorithms,three kinds of emotion classification models at specific document level have been studied.Then,according to the different classification model,each model is implemented in detail.Based on the study,the experimental results of each model were analyzed in detail.This paper introduces the research background,significance and current situation of Cambodian news commentary emotion emotion classification at home and abroad,and the existing problems.Secondly,based on the emotional classification of sentence level,three kinds of specific emotion classification models are summarized,which provide some ideas for the establishment of the thesaurus and the construction of the experimental platform.On the basis of the document level,the reasoning and the algorithm of document level are introduced and studied respectively.Based on the optimization of the basic algorithm of document level,three kinds of specific emotion classification models are given.The construction of the news corpus in Cambodia,the construction of the experimental platform,the analysis of the experimental results,and the performance of several models were compared.This paper describes the design of the emotion classification system,the results of the analysis of the emotion classification system,introduces the performance index,gives the experimental results of the experimental model,and gives a comparison of the performance of each model to discuss the impact of each model the elements of.Finally,the work of this paper is summarized and the follow-up research related to the subject of this paper is expected.