摘要：

Abstract:

【Purpose/significance】 Understanding the textual characteristics of user-generated content in different types of social me⁃

dia platforms, exploring the difference of topic shift and content extension, has important theoretical value and practical significance to

better guide user-generated content

.

【Method/process】 The research classifies the main content forms released by the platform, selects

four social media platforms, and uses keyword mining, word co-occurrence, and sentiment analysis techniques to compare and analyze

the different language style, emotional tendency, topic shift, and content extension of user-generated content.【Result/conclusion】 The

results show that in terms of textual features, different platforms have their own unique language content features and language styles.

In terms of theme and content, there are differences in the degree of shift and content of topic shift and content extension phenomenon

on different platforms.【Innovation/limitation】 Based on the common types of social media content presentation, this study innovatively

explores the different textual characteristics, topic shift, and content extension of user-generated content on different types of social

media platforms from a cross-platform perspective. In the future, the user-generated content from multiple topic events can be com⁃

bined to further enrich the research methods and explore the reasons behind the differences.