Managing multilingual collections: insights from data analytics research

Libraries, through their catalogues and borrowing records are well-placed to use data analytics to enhance their collection management (and of course do this already, for example by directing orders to genres/areas that are heavily borrowed). In this article, we explore some of the insights for the management of multilingual collections offered by a novel research method that fuses analysis of a large data set of borrowing records with data from interviews with library staff. Such a method, we argue, helps to untangle the Gordian knot around why materials in some languages are widely popular while materials for other equally widely-spoken languages sit unused on the shelves. It also draws our attention to the ways in which different demographics of speakers are engaging with library materials across the various languages, and gives a suite of tools local libraries might use to better assess the likely demand for materials in languages other than English.