FuzzyPhoto

The FuzzyPhoto project was a two year AHRC funded research project (AH/J004367/1) to develop computational methods for finding the pictures "missing" from the catalogues of the annual exhibitions of the Royal Photographic Society 1870-1915 by matching the textual exhibit records with photographic item records in a range of partner institutions. Project partners awee Birmingham Central Library, the British Library, the Musée D’Dorsay and the Louvre, Paris, the Metropolitan Museum of Art, New York, the National Media Museum, St Andrews University and the V&A. The project started in November 2012, ending in November 2014. The particular challenges it tackled, in addition to the usual problems of messy data, were very large numbers (>1.4 million) of very short records (on average only 5.4 words per record). To deal with these a new Semantic Short Text Similarity Metric was developed that outperforms previous tools of this type.