Online Perception Model of College Students' Emotion Under the Fusion of Social Media Such as Weibo

Emotional analysis of netizens' expressions on social media is a branch of natural language processing that aims to mine people's opinions and emotional tendencies toward events and things. Analyzing public opinion in higher education has become a trend in recent years, helping guide the reform of higher education. This paper presents an optimized design scheme for college pupils' emotional online perception model using Weibo comments. Three types of emotional articles—“joy,” “anger,” and “sadness”—are collected and preprocessed. Word2Vec is used for algorithmic network feature extraction, and key features are annotated with emotions in an artificial dictionary. Hidden themes or topic categories are found from a large number of documents, and emotion classification is completed assuming each document is a mixture of multiple topics. The study of online emotional perception in colleges has high practical significance and value, aiding in the implementation of control measures and psychological development of pupils.