Social media analytics in museums: extracting expressions of inspiration

ABSTRACT

Museums have a remit to inspire visitors. However, inspiration is a complex, subjective construct and analyses of inspiration are often laborious. Increased use of social media by museums and visitors may provide new opportunities to collect evidence of inspiration more efficiently. This research investigates the feasibility of a system based on knowledge patterns from FrameNet – a lexicon structured around models of typical experiences – to extract expressions of inspiration from social media.

The study balanced interpretation of inspiration by museum staff and computational processing of Twitter data. This balance was achieved by using prototype tools to change a museum’s Information Systems in ways that both enabled the potential of new, social-media-based information sources to be assessed, and which caused the museum staff to reflect upon the nature of inspiration and its role in the relationships between the museum and its visitors. The prototype tools collected and helped analyse Twitter data related to two events. Working with museum experts, the value of finding expressions of inspiration in Tweets was explored and an evaluation using annotated content achieved an F-measure of 0.46, indicating that social media may have some potential as a source of valuable information for museums, though this depends heavily upon how annotation exercises are conducted. These findings are discussed along with the wider implications of the role of social media in museums.

KEYWORDS:

• Museum
• inspiration
• social media
• analytics
• semantic analysis
• natural language processing

Acknowledgements

Many thanks to Emma Hallam (Social Media Coordinator) and Jonathan Wallis (Head of Museums) at Derby Museums, and to Derby Museums as a whole. This work would not have been possible without the willingness to experiment which is fundamental to their organisation.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Dr. David Gerrard is the Digital Preservation Technical Specialist (Polonsky Fellow) at Cambridge University Library, researching ways to incorporate the preservation of digital materials into the Library’s business processes. He received a PhD from Loughborough University in 2016, after researching the potential for museums to use information from social media to discover more about their visitors.

Dr. Martin Sykora is a Lecturer with main research interests in the areas of social-media and big-data analytics, with a significant track record and expertise in the sentiment analysis, text mining and machine learning fields. Martin’s work on sentiment analysis has been applied to public mental health surveillance after the Paris 2015 terror attacks, with a short paper recently accepted for publication in the Lancet.

Prof. Thomas W. Jackson (BSc, PhD, FBCS) is a Professor of Information and Knowledge Management and is the Director of the Centre for Information Management, and the Associate Dean for Research in the School of Business and Economics at Loughborough University. He has over 15 years’ experience of research and industrial consultancy. His research areas are Electronic Communication and Information Retrieval, and Applied and Theory based Knowledge Management, including Natural Language Processing.

ORCID

David Gerrard http://orcid.org/0000-0002-5659-4720

Martin Sykora http://orcid.org/0000-0002-5363-5857

Thomas Jackson http://orcid.org/0000-0001-8404-4251

Additional information

Funding

This research was partly funded by the UK Arts and Humanities Research Council [grant number 1234317].