Perception of Cross-cultural Garden Tourism: A Case Study of the New York Chinese Scholar's Garden at Snug Harbor
This study analyzes textual reviews and photographs of the New York Chinese Scholar's Garden to reveal that while architecture and vegetation are common themes in both media, the visual composition of visitor photos and their verbal sentiment evaluations are largely independent, with visual elements explaining less than 7% of the variance in review sentiment.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
When people travel to a place that feels different from home, they often try to capture the experience in words and pictures. They might write a review describing how peaceful a garden feels, or snap a photo of a stone bridge over a koi pond. For decades, researchers have studied how these digital records shape our understanding of a destination, treating the collection of online posts as a kind of collective memory. At the same time, other scholars have looked closely at the physical design of gardens, analyzing how architects arrange rocks, water, and plants to create a specific mood. What has remained unclear is how these two ways of seeing the same place—the written description and the visual snapshot—actually relate to one another. Do the things people photograph most often match the things they praise in their reviews? Does a picture filled with green trees guarantee a happy review, or are the two records telling different stories about the same visit?
To answer these questions, a team of researchers turned their attention to the Chinese Scholar's Garden at Snug Harbor in Staten Island, New York. Built in the style of classical gardens from Suzhou, China, this space offers a unique setting where traditional Chinese design meets the everyday tourism of the United States. The researchers gathered a massive amount of data from three popular travel platforms: Yelp, Google Maps, and Twitter. They collected nearly 1,400 written reviews and over 1,300 photographs posted by visitors between 2008 and 2025. The goal was not just to count how many people liked the garden, but to see if the visual content of the photos could predict the emotional tone of the written reviews. They wanted to know if the camera and the pen were recording the same experience or two entirely different ones.
The team first looked at what visitors were actually saying. After filtering out irrelevant posts, they analyzed 942 reviews. The results showed that most visitors felt positively about the garden, with nearly 58 percent of the reviews expressing happiness or satisfaction. The words that appeared most often described the place as beautiful, peaceful, and authentic. When the researchers broke down the specific garden features mentioned in the text, architecture came out on top. People frequently wrote about buildings, pavilions, and roofs, followed closely by mentions of plants, particularly bamboo. Water features and rocks were also discussed, but less often than the built structures and the greenery. This suggested that when visitors put their thoughts into words, they were drawn to the human-made elements of the garden as much as, or perhaps more than, the natural ones.
Next, the researchers turned their attention to the photographs. Using computer software capable of recognizing objects in images, they analyzed 1,395 photos to see what elements appeared most frequently. The results here were slightly different from the text. The most common things captured in the photos were trees, plants, and the sky. While buildings and rocks also appeared often, the natural elements dominated the visual record. The software also looked at how these elements appeared together. It found that the most frequent combination in a single photo was a plant next to a building. This pairing happened in nearly 900 photos, far more than any other combination. Water and rocks often appeared together as well, but they were less likely to be the sole focus of an image. The visual data painted a picture of a garden where nature and architecture are almost always seen side by side, framing the visitor's view.
The most revealing part of the study came when the researchers tried to connect the two sets of data. They looked at the specific photos that were posted alongside written reviews and asked a simple question: could the things visible in the photo predict how positive or negative the written review would be? They used several different computer models to test this relationship, treating the presence of elements like trees, water, or buildings as clues to guess the visitor's mood. The answer was surprising. The models failed to find a strong link. No matter how they adjusted the analysis, the visual composition of a photo could not reliably predict the sentiment of the accompanying text. The best model managed to explain less than 7 percent of the variation in how people felt.
This finding suggests that the way people see a garden with their eyes and the way they describe it with words are largely independent acts. A photo might be filled with lush greenery and a clear blue sky, yet the person who took it might write a neutral or even critical review about the service or the crowd. Conversely, a photo of a simple stone wall might be paired with a glowing review about the garden's tranquility. The researchers concluded that for overseas Chinese gardens, the visual traces left by visitors and the verbal descriptions they write are complementary records, not interchangeable ones. They capture different facets of the experience that do not necessarily align. The study does not prove that one way of recording is better than the other, but it does show that we cannot assume a beautiful picture guarantees a happy story, or that a happy story requires a picture of a specific landmark. In the end, the digital footprint of a visitor is a complex mix of what they chose to show and what they chose to say, and the two do not always tell the same tale.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.