• Computation and Rhetorical Invention: Finding Things to Say With word2vec

    Author(s):
    Mark B. Wolff (see profile)
    Date:
    2019
    Group(s):
    DH2020, Digital Humanists, Digital Pedagogy
    Subject(s):
    Computational linguistics, Literature--Computer-assisted instruction, Machine learning
    Item Type:
    Abstract
    Tag(s):
    rhetorical invention, word2vec, Computational lingustics, Computers in literary instruction, Electronic literature
    Permanent URL:
    http://dx.doi.org/10.17613/7th3-1396
    Abstract:
    Macroanalysis and machine learning do not only lend themselves to interpreting literature: they can be used to generate literature within the rhetorical affordances of digital technology. With computation, our ability to access and query texts has changed our relationship to them. Engaging digital texts as big data allows one to explore rhetorical invention, or the finding things to say with language. David Johnson’s ReRites (2018-19) is an example of poetry created in large part with algorithms: a neural net selects text from a large corpus and the poet edits the output. Another example is my sonnet generator, which uses large corpora of poetry in French, English and Spanish to enable a user to select verses semantically using word2vec and assemble a new poem (with the possibility to alter the text). Computational literary studies can extend beyond literary analysis and synthesize new texts using algorithms as inventional heuristics.
    Notes:
    A demo of the code for generating sonnets is available here: http://markwolff.name/wp/digital-humanities-2/invent-your-own-sonnet-using-analytic-tools-to-synthesize-texts/
    Metadata:
    Status:
    Published
    Last Updated:
    4 years ago
    License:
    Attribution

    Downloads

    Item Name: pdf dh2020-wolff-abstract.pdf
      Download View in browser
    Activity: Downloads: 111