Sentiment Analysis Of Weibo Information Based On The Combination Of Ontology And Syntactic Dependence

With the popularity of the mobile Internet,Weibo has become the fastest source of information outbreaks as a social platform.For enterprises,the emotional information contained in Weibo user blog posts has become an important decision-making basis for enterprises.In order to obtain the emotional information from Weibo easily and accurately,a microblog sentiment analysis method based on the combination of ontology and syntax dependence is proposed.The blank ontology containing the core words is persisted to the database,and then the syntactic dependency is extracted from the collected microblog data to update the ontology using the syntactic dependency principle.Then a new emotional weight calculation method is proposed,which will be the ontology dimension.Information analysis takes into account emotional analysis.Experiments show that this method can effectively determine the emotional tendency of Weibo data.Next is the most critical part of this paper:1.Use the associated nouns of the core words to expand the corpus.In the evaluation data of netizens' microblogs,core words and sensitive words are sometimes omitted,and the products are also referred to as the products.The information items included in the proxy are sometimes even more than the original ones.Therefore,the analysis using only the original name is not objective and comprehensive..In this paper,the original data is reduced by the tool,and the nouns that are most similar to the original name are obtained from the original data,and then the corresponding microblog data is crawled to achieve the purpose of expanding the data.2.Use the syntactic dependency to expand the dimensions of the ontology.The ontology relies on manual creation.This method covers a situation where the dimension is small and the workload is huge,which will have a great impact on efficiency,and the dimension will have an empty dimension.In this paper,we use the syntactic dependency relationship to discover the dependency relationship related to the core ontology from the microblog data,and then use the related words of the dependency relationship to automatically expand the ontology dimension.3.Use the syntactic dependency to expand the ontological emotional vocabulary.There are some specific emotional words in the dimension of the ontology,which can't be expected at the time of initial creation.For the words that need to be mined in the data,this paper uses the syntactic dependency to find the corresponding emotion in the traversal of the data.Vocabulary,and added to the ontology system,to achieve the purpose of automatically expanding emotional vocabulary.4.A new calculation method is proposed to make use of text sentiment weights.This paper proposes an emotion weight analysis method based on ontology dimension,which takes into account the ontology dimension of sentiment words,so that the spatial position information of emotional words can be fully utilized to avoid the situation that unimportant words represent excessive weights.Appears to more fully reflect the emotional sentiment of the data as a whole.5.Taking Mengniu as the research object,using tools to complete the crawling of Mengniu microblog data,discover the core nouns associated with the nouns,then crawl the related nouns and filter them,and then use the data to complete the "Mengniu" ontology.Persistence and content expansion,and finally calculate the emotional orientation.Through experimental comparison results,it is proved that the new method constructed in this paper has a long-term advantage over other methods in depicting the emotional tendency of Weibo data..