Tourist Engagement Analysis at Gao Family Compound via Multimodal Social Media Data

6.2 GB of multimodal data applies BERT-based sentiment analysis and a fine-tuned ViT-Large model across 31 visual classes to social media posts from Weibo and Xiaohongshu. The study by Liyue Wu examines how tourists perceive the Gao Family Compound in Xi'an, linking visual attention with affective attachment. The dataset was last updated on 2026-05-24.

Use Cases

• Analyze tourist sentiment polarity based on BERT-processed social media text.
• Classify tourist images of heritage sites based on the 31 architectural and cultural visual classes.
• Identify management pain points based on negative posts related to service and pricing.
• Study cross-modal alignment between textual keywords and visual attention patterns.

Strengths

• Multimodal approach integrates text and image data from two major Chinese social media platforms.
• Fine-tuned ViT-Large model provides classification across 31 specific architectural and cultural classes.
• Results include specific findings like a predominance of positive sentiment and focus on symbolic forms.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect geographic and platform bias inherent to Weibo and Xiaohongshu.

Provenance

Source

Liyue Wu via figshare.

Collection Method

Social-media texts and images collected from Weibo and Xiaohongshu, processed with multimodal deep learning pipelines.

Freshness

Last updated 2026-05-24 13:34:23; freshness should be verified.

Geography

Gao Family Compound, Xi'an, China.