Research On The Construction Of Sentiment Analysis Model Of Long Text Driven By Reading Context

At present,the reading platform has added the context to the reading.Through investigation,it is found that the current context is displayed in an audio-visual manner,but these reading contexts have nothing to do with the emotion in the text.In order to add emotion to the audio-visual context of the reading platform.It is necessary to extract the emotions in the text,and match the obtained emotions with the pictures and sounds in the database,so as to construct an emotional situation.This article mainly focuses on the acquisition of text emotions,using emotion dictionary and semantic rules to analyze the text,but the emotions obtained through the above methods lack the text background,which leads to the inability to accurately match the corresponding pictures and music when constructing the audiovisual reading situation.In order to solve the above problems,this article adds emotion summarization technology to text sentiment analysis.Finally,a sentiment analysis model for long texts driven by reading context is constructed.The thesis consists of six parts: The first part is an introduction,mainly describing the research background and significance,and researching the current situation at home and abroad,and introducing the research methods and content of this article.The second part is about the related theories and the technology of sentiment analysis Introduction.The third part verifies the relationship between emotion and reading context and the user's needs for emotion when reading through demand research.The fourth part uses sentiment dictionary and semantic rules to analyze text to obtain emotions,and uses emotion summary technology to obtain text emotion topics,constructing a reading context-driven long text sentiment analysis model.The fifth part conducts empirical research on the constructed model.The sixth part makes suggestions for the future development of the model.The text sentiment analysis model constructed in this paper is constructed from the reading context,which has a positive effect on constructing an emotional audio-visual reading context for the reading platform.