.png){kind=icon}
.png){kind=icon}
Return
Evaluating cultural heritage value through multi-source data fusion and sentiment analysis: a case study of 74 heritage sites
.png){kind=icon}
OA
AI
X
G
Y
Y
T
C
DOI：10.1038/s40494-026-02588-3.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Accurate evaluation of cultural heritage value is crucial for its preservation and sustainable development. However, previous studies have largely relied on single-source data. This study assesses 74 heritage sites in Ningbo, China, by integrating 24,210 visitor reviews with objective data. We applied the RAKE algorithm to extract keywords and assign dynamic weights to four objective dimensions: history, culture, protection, and transport. Fine-grained sentiment analysis was conducted using Paddle NLP, with Bayesian smoothing to mitigate small-sample bias. Key findings include: (1) over 60% of objective weight was attributed to culture and history; (2) heritage value was primarily shaped by objective attributes, with weak ties to sentiment, revealing resource-satisfaction imbalances; and (3) higher evaluations were associated with better-protected and centrally located sites. This study demonstrates that text mining and sentiment analysis offer effective means for heritage assessment, and introduces an attribute–perception–experience framework to clarify how social perceptions shape heritage value.
Keywords:
cultural heritage
text mining
sentiment analysis
multi-source data fusion
heritage value
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.