Identification and Visualization System Construction of Collective Memory Elements in Traditional Villages Based on Multi-source Tourist-Generated Content (UGC) and Knowledge Graphs

This study proposes an integrated methodological framework that combines multi-source tourist-generated content analysis with knowledge graph construction to identify and visualize the constituent elements of collective memory in traditional villages. In the context of increasingly intelligent cultural communication supported by electromagnetic information infrastructures, wireless sensing networks, and digital interaction technologies, the framework specifically explores how traditional craftsmanship and heritage function as material and symbolic anchors of cultural authenticity in village conservation. To mitigate the demographic bias inherent in social media data, a dual-weighted stratified sampling strategy is adopted to systematically incorporate indigenous perspectives. Multi-dimensional information collected from social media, professional literature, and field investigations is processed through fine-tuned pre-trained language models for semantic mining and entity extraction, while a “Contradiction” entity class is introduced to preserve conflicting narratives between tourists and residents. Building upon a multi-level ontology encompassing material spaces, non-material practices, and emotional cognition, the proposed knowledge graph structurally represents complex memory associations and supports dynamic expansion. An interactive visualization system is further developed and validated with target users, achieving an 86.7% satisfaction rate by effectively tracing the recent evolution of traditional motifs and artisanal techniques in tourism discourse. The results demonstrate that the proposed framework provides an interpretable and scalable digital solution for cultural heritage communication over modern information transmission environments while strengthening the preservation and dissemination of collective memory.