Museums often struggle to find ways of evaluating visitor experience without the use of intrusive instruments such as exit interviews and surveys. In the last decade, visitors have increasingly used social media to post reviews, photos, and comments about their experiences to their networks. This data represents a relatively untapped opportunity for museums to understand what engages their visitors, and how their experiences conform to or challenge expectations. This project will pilot the use of social media data to enhance our understanding of ‘playful’ visitor behaviour across Oxford’s GLAM. Using data from Twitter and Instagram, we will look for new types of engagement, with a view to trying to understand how visitors experience Oxford’s collections as ‘playful spaces’.

The project will produce a dataset of social media posts about Oxford’s GLAM collections. It will also produce a methodology for analysing these data, by applying natural language processing and image classification algorithms, among other machine learning tools, to cluster similar image types together, and reveal the most common themes in pictures posted online by GLAM visitors. Computational data analysis, and pattern recognition tools such as these, will allow us to tackle the volume and diversity of content posted to social media websites, in order to uncover common patterns of engagement within the collections.

We will then explore the idea of co-creating a small number of in-gallery experiments to share our findings and to encourage visitors who wish to participate to experience the museums in a playful manner.  We will also use the data to create a community-generated map of playful spaces in Oxford’s gardens, libraries and museums.