Linking Motif Sequences with Tale Types by Machine Learning

摘要

Abstract units of narrative content called motifs constitute sequences, also known as tale types. However whereas the dependency of tale types on the constituent motifs is clear, the strength of their bond has not been measured this far. Based on the observation that differences between such motif sequences are reminiscent of nucleotide and chromosome mutations in genetics, i.e., constitute "narrative DNA", we used sequence mining methods from bioinformatics to learn more about the nature of tale types as a corpus. 94% of the Aarne-Thompson-Uther catalogue (2249 tale types in 7050 variants) was listed as individual motif strings based on the Thompson Motif Index, and scanned for similar subsequences. Next, using machine learning algorithms, we built and evaluated a classifier which predicts the tale type of a new motif sequence. Our findings indicate that, due to the size of the available samples, the classification model was best able to predict magic tales, novelles and jokes.