Research on the“Semantic Reversal”Mechanism and Precise Governance of Dialect-Based Public Sentiment in a Multimodal Context:A Cross-Platform Analysis Based on Chongqing's“Fat Cat Incident”

Abstract

Addressing the challenges of modal fragmentation and dialect-based“semantic divides”in multimodal online public opinion governance, this study employs the Chongqing“Fat Cat Incident”as an empirical case to construct a dialect-context-oriented cross-modal semantic alignment model (D-CMSA).Through cross-platform comparative analysis of 8,542 Weibo posts and 312 high-engagement Douyin videos,we measured the emotional evolution trajectories of dialect vocabulary and the mobilization mechanisms of visual symbols.Empirical validation confirms that the D-CMSA model possesses the capability to uncover deep public sentiment features difficult to capture through single-modality analysis.These features can be distilled into two core effects:First, the sentiment amplification effect,where the co-occurrence of negative audiovisual symbols increases negative sentiment intensity by an average of 42.9%. Second,the semantic reversal effect:28.5% of samples demonstrate that dialect texts with neutral surface semantics are reconfigured into ironic or resentful symbols within specific visual contexts,exhibiting significant phased semantic drift characteristics.This study proposes establishing a dynamic dialect sentiment lexicon to address semantic drift and employs cross-modal consistency verification algorithms to prevent covert dissemination in public opinion governance.