MCN 2024

Exploring AI and machine learning in art museums often feels like an exercise in separating hype, achievable near-term value, and potential long-term game-changers. In this presentation, the National Gallery of Art will share how that pursuit is playing out after 18 months of pilots with cross-functional teams in two priority use cases, with lessons learned to date and plans for the way ahead. First: after initial work to scan, run optical character recognition (OCR), and analyze exhibition response wall cards and visitor comments, the team found that an AI-powered chatbot built within the network helped quickly find insights among thousands of comments, unlocking new value from qualitative data. Second: as a part of ongoing transformation in exhibition planning and operations, machine learning helped mine a decade of data to predict attendance curves and gauge what drives audience engagement. The data science team will present data visualizations, predictive modeling techniques, and methods for natural language processing and chatbot development, while members of visitor experience and evaluation will share findings, time savings, and future plans from these two initiatives. Recognizing that the value of analytics projects is measured by the decisions and outcomes they inform, the session will address how the results are used and future plans for plugging into business processes, with relevance to any museum and an invitation to participate in ongoing analysis, benchmarking, and collaborative data culture across museums.