Urban public spaces are constantly restructured with new meaning, reflecting their socio-cultural, political, historical, and technological influences. Over the last two decades, the rapid technological advancements and increasingly widespread use of mobile devices give people a chance to share their experiences of their immediate surroundings through various applications. As these platforms enable people to create and exchange various forms of User-Generated Content (UGC) has gained wide attention as an invaluable source of information on human-environment relationships including people's timely perception, emotion, preference, and sense of place in public space.

This study employs quantitative content analysis to identify collective perceptions of urban public spaces and their characteristics as projected through a photo-sharing social media application, Instagram. A total of 1,200 users' photos and associated hashtags geo-referenced to three New York City urban public spaces, Bryant Park, Madison Square Park, and Union Square. This study begins with a qualitative phase, employing manual categorization techniques to identify the concepts in visual and textual data. The second phase applied a statistical analysis method, a set of descriptive analytics, and chi-square tests to answer the research questions for this study.

Findings indicate physical attributes of urban squares are the most dominant type of geo-referenced users' photographs through the visual social media platform. People's immediate perceptions vary with time and place, while the patterns of hashtag usage found in this study show no difference across the three urban squares. people's perception of urban squares goes beyond the boundaries of the square itself, encompassing the streetscape, buildings, and local businesses adjacent to the square. While people rarely utilize hashtags as a method of emotional expression, findings show a clear connection between hashtags associated with users' photo content and the image of the city.