Research on the Optimization of Translation and Dissemination Strategies for Central Plains Culture and Overseas Publicity Based on Audience Analysis

Central Plains culture faces a persistent cultural discount problem in overseas translation and dissemination. Traditional publicity translation is mainly text-centered and often lacks quantitative analysis of audience expectations, producing a mismatch between translated content and audience cognition. In digital and wireless media environments where cultural information is transmitted through multimodal channels, audience-oriented modeling is essential for improving communication efficiency. This study develops a closed-loop optimization model consisting of audience profiling, strategy adjustment, and effect feedback. Web crawling and LDA topic modeling were first used to classify overseas audiences into curiosity seekers, explorers, and audiences with cultural misconceptions according to topic probability thresholds and sentiment polarity. A layered progressive-disclosure text structure was then designed so that readers could choose either reconstructed cultural frames or deeper explanatory annotations. Finally, eye-tracking, reading comprehension tests, and focus group semantic analysis were used for multidimensional validation. The optimized translation reduced average gaze duration by 73.2 ms, decreased return gaze rate by 8.0%, increased message comprehension by 19.5 percentage points, and improved the proportion of positive emotional vocabulary by 18.4 percentage points. The findings demonstrate that data-driven translation design can reduce cognitive load and enhance cultural identification in international communication.