弹幕文本挖掘与情感分析

Danmu Text Mining and Emotion Analysis

摘要: 弹幕文本数据的流行，为短文本处理和实时数据处理提供了大量新的文本数据。本文首先对近年来关于弹幕文本的研究进行了系统性梳理归纳，然后基于文本挖掘技术对节目视频弹幕进行深层数据分析，围绕弹幕文本情感分析的关键技术和基本流程进行重点阐述，主要包括通过Python进行文本获取、文本预处理、高频词与词云图可视化、弹幕文本主题词分析、弹幕文本情感分析等多个模块，完成弹幕情感倾向分析，探究弹幕文本数据结构及文本特征，提高弹幕文本情感分析准确度。

Abstract: The popularity of bullet screen text data provides a lot of new text data for short text processing and real-time data processing. Firstly, this paper systematically summarizes the researches on bullet screen text in recent years, and then analyzes the deep data of program video bullet screen based on text mining technology. This paper focuses on the key technologies and basic processes of bullet-screen text emotion analysis, including text acquisition, text preprocessing, visualization of high-frequency words and word cloud images, bullet-screen text theme word analysis, bullet-screen text emotion analysis and other modules through Python to complete bullet-screen emotion tendency analysis and explore the data structure and text characteristics of bullet-screen text. Improve the accuracy of emotion analysis of bullet screen text.