Discovering Multilingual Text Reuse in Literary Texts

We present here a method for automatically discovering several classes of text reuse across different languages, from the most similar (translations) to the most oblique (literary allusions). Allusions are an important subclass of reuse because they involve the appropriation of isolated words and phrases within otherwise unrelated sentences, so that traditional methods of identifying reuse including topical similarity and translation models do not apply. To evaluate this work we have created (and publicly released) a test set of literary allusions between John Milton’s Paradise Lost and Vergil’s Aeneid; we find that while the baseline discovery of translations (55.0% F-measure) far surpasses the discovery of allusions (4.8%), its ability to expedite the traditional work of humanities scholars makes it a line of research strongly worth pursuing.