Research On Patriotic Feelings Of Classical Poems Based On TextCNN

Classical poetry has always played an important role in Chinese literature and has been regarded as a gem of Chinese literature and culture.Therefore,classical poetry has a very high research value and profound research significance..However,there are countless poetry works handed down in China over the past millennia.According to traditional research methods,researchers need to go through a large number of reading reserves,and then read,search,record,organize,analyze,and count on many literary works.Personal judgment.In this process,researchers not only need to invest a lot of manpower,material resources,energy,and time,but also have high requirements for researchers' literary literacy.They will also make unobjective conclusions due to personal preferences.The country's constant advocacy of the core values of socialism shows that the patriotism is very important to everyone living in a fast-developing modern era.Patriotic feelings contain a long history and culture,and ancient poetry has always been a literary form that Chinese people love very much.It is very helpful to inspire people's patriotic feelings by propagating and promoting patriotic poetry.The study of patriotic poetry will help spread Chinese culture,promote the national spirit,and cultivate core values of socialism.This paper uses the statistical model of text mining and computer program language to consider the characteristics of the literary form of ancient poetry in the study,classify classical poetry and classify it into patriotic and other categories.After determining the classification model,use this The model classifies more and more comprehensive data of ancient poetry to obtain patriotic poetry,and then analyzes the patriotic poetry in general to explore the patriotic feelings in ancient poetry.The classification methods used are Text CNN,a convolutional neural network that is specifically oriented to text data in deep learning,and SVM,a classic classification model in machine learning,and introduces word vectors for text representation.The research work of this paper mainly includes the following aspects: 1.This article summarizes the current development of text mining and the use of statistical methods and computer knowledge to study the status of classical poetry;2.This paper introduces the word vector as a distributed method for text representation of classical poems.Compared with the traditional bag-of-words model,it avoids the problems of dimensional disaster and sparse features.Moreover,the word vector takes into account similar words and synonyms,and more Able to express the refined classical poetry text;3.Classification use SVM and Text CNN to construct a classification model.By comparing the classification effects,a model suitable for classical poetry classification is finally determined;4.Applying the determined Text CNN classification model to more classical poetry data,we have obtained a relatively complete text data of patriotic poetry covering different authors in different periods,an overall analysis of it,and the famous patriotic poet Lu You's patriotic poetry As an example,analyze the frequency of words to study the patriotic feelings contained in ancient poetry.Through these aspects of work,each ancient poem has been converted into a set of 200-dimensional data,and whether it is a patriotic poem can be judged through a classification model.The overall analysis of patriotic poems can be found in different historical periods The degree of patriotism is also different.By analyzing the frequency of words,you can better understand the patriotism in classical poetry.