The mixed-methods approach examined colour schemes, compositional patterns and symbolic codes in shared images alongside automated sentiment classification of accompanying text. Hamamönü emerged as a neighbourhood defined primarily by traditional architecture, which accounted for 41 per cent of its representation, followed by gastronomy-focused experiences at 34 per cent. This distribution suggests that the district’s digital image centres on its cultural and experiential identity rather than on any single monument.

Ankara Castle presented a different profile. Archaeological heritage dominated its social media portrayal at 38 per cent, with the urban landscape contributing a further 29 per cent, underscoring the site’s historical and spatial dimensions as perceived by online audiences. The Museum of Anatolian Civilizations, by contrast, was most strongly associated with educational value, cited in 45 per cent of representations, and with interactive exhibition practices at 27 per cent, reflecting its prominent role in digitally mediated museum experiences.

Sentiment analysis revealed a clear polarization across the three sites. Museum-related posts were overwhelmingly positive, with 82 per cent classified as favourable, while criticism concentrated elsewhere: 28 per cent of negative commentary targeted Ankara Castle’s infrastructure and 19 per cent concerned visitor management issues. The authors argue that this pattern demonstrates how social media analytics can guide evidence-based strategies for digital transformation, promotion and sustainable communication of cultural heritage sites.

The study offers a data-driven framework for heritage managers seeking to understand how digital transformation reshapes the representation and perception of cultural heritage within contemporary tourism, showing that the emotional tone and visual vocabulary of online posts can be systematically measured rather than merely assumed.

Subject of Research: Semiotic and NLP analysis of social media representation of cultural heritage sites in Ankara

Article Title: Between pixels and emotions: semiotic-NLP deconstruction of cultural heritage’s social media image

Article References: Çetin, Y., Islak, Y., Çınar, E., Ertürk, M., & Hassan, A. (2026). Between pixels and emotions: semiotic-NLP deconstruction of cultural heritage’s social media image. Humanities and Social Sciences Communications. https://doi.org/10.1057/s41599-026-09538-1

Image Credits: AI Generated

DOI: 10.1057/s41599-026-09538-1

Keywords: Between, pixels, emotions, semiotic-NLP, deconstruction, cultural, heritage, social, media, image, scientific research