Utilizing Metadata from Heterogeneous Sources within the Framework of the JVMG and GOLEM Projects to Identify Patterns in Anime-based Fandoms on AO3

This paper offers an overview of the first step in the data alignment and joint research efforts between the Japanese Visual Media Graph (JVMG) and the Graphs and Ontologies for Literary Evolution Models (GOLEM) projects. Focusing on fully or in-part anime based fandoms in the Archive of Our Own (AO3) data in the GOLEM knowledge graph we connected 534 fandoms with 1,734 anime works from the JVMG knowledge graph. The fanfiction publication numbers and dates for these fandoms were then analyzed in conjunction with the various metadata available on the corresponding anime works, such as release date, genre, content rating and user rating. Fixed effects modeling was employed to identify a positive effect of anime releases on fanfiction publication numbers. Furthermore, individual graphs of the top four hundred fandoms’ fanfiction publication numbers over time overlayed with the corresponding anime release dates were also analyzed manually to identify a number of distinct patterns, as well as groups of fandoms for which the invigorating effect of anime releases on fanfiction publication numbers could be clearly identified.