The purpose of this paper is to analyze the influence of Milton’s Paradise Lost on a total of 292 of Coleridge’s poems and dramatic works, using a quantitative method to visualize the cosine similarities between Paradise Lost and Coleridge’s works. The paper adopts two similarity models of so-called “Paradise Lostness” and “Whole Paradise Lost.” Paradise Lostness, based on Piper and Algee-Hewitt’s (2014) “Wertherness,” uses the top 100 most frequent words in Paradise Lost divided by the total words in each poem, and computes the cosine similarity among them. Whole Paradise Lost is used to compute the cosine similarity using the tf-idf weighting scheme.

  The result of the first attempt showed that poems with higher cosine similarities in both models tend to have more words. Therefore the polynomial regressions were then computed in two graphs to diminish the effect of the number of words on cosine similarity. From those two graphs, poems with higher cosine similarity than the standard deviation 1 of the polynomial regression line were extracted.

  Judging from the close reading of those extracted poems, the paper found that there are latent similarities between Paradise Lost and several of Coleridge’s texts. Poems such as “Ode to the Departing Year,” “Hymn before Sun-rise, ㏌ the Vale of Chamouni,” “My Baptismal Birth-Day,” “The Improvisatore,” and “Lines Suggested by the Last Words of Berengarius Ob. Anno Dom. 1088” are similar to Paradise Lost, as well as “Ne Plus Ultra” and “Religious Musings”, as previous critics have pointed out.

  On the other hand, poems such as “The Rime of the Ancient Mariner” and “The Destiny of Nations,” for which critics have shown the influence of Paradise Lost, are not plausibly similar to Paradise Lost, possibly because of its indirect influence and the multitude of different topics the poem describes.

  The paper concludes that the visualizing method of similarity measure among Paradise Lost and Coleridge’s works showed that several newly found poems have a latent similarity, and that some poems critics have found to be similar are not plausibly similar. Although it has several problems, the paper proposed one of the quantitative methods for the study of influence of Romantic texts.