Social media data mining for exploring readers’ literary interests

Abstract

Digital humanities have become a booming field of study thanks to the application of computer and information science methods to the humanities and the accumulation of large-scale digital resources. The use of data mining techniques and quantitative methods is expanding and starting to dominate the domain of literary studies, focusing not only on the study of the text itself but also on the study of social influences and preferences of the readers. By analysing social network data from VK, the most popular Russian platform, the study investigates which authors and literary works resonate most with users. To analyse the quotations, we created a framework which combines tools for extracting and cleaning data, named entity recognition and finally, corpus analysis; this framework can be applied to other datasets to compare and expand our findings. We also noted that within the information behaviour research domain, such studies of sharing among communities are not yet very popular. The study showcases the importance of employing advanced computational tools in humanities research. The obtained quantitative results are subjected to a critical analysis, which can serve as a basis for humanitarian understanding based on data. The compelling findings of this pilot study confirm opportunities for further research using more advanced quantitative methodologies, as well as broadening the scope of criteria that potentially influence the formation of reading preferences, such as level of education, pop-culture trends, social environment, etc. Additionally, it sheds light on the underexplored domain of community building in social media networks, calling for broader research into information behaviour and online community dynamics.