Posts & Files & Varia

Stylometry, Saga Corpus & Egils saga

• Saga Corpus: Link to the data-set (by Eiríkur Rögnvaldsson & Sigrún Helgadóttir, 2011/2022) at Clarin-IS. If the download issue has not been resolved yet, see: repository.clarin.is.

• Saga Corpus MDS 3-in-1: Saga Corpus 1-, 2-, 3-gram 1000 MFW/MFF visualized with Multidimensional Scaling. Interactive plotly plot.

• Video: 10-1000 MFW/MFF centered around 5 sagas from SC in order to hypothesize about the influence of the size of the features on authorship conjectures. Using Multidimensional Scaling for dimension reduction.

• Video: Same as above, but includes all texts.

• Different plots based on Burrows’ Delta distance measures(in PDFs):

  • All of the bellow and some more: All-in-one
  • PCA 1000 MFW/MFF:
    • PCA 1-gram & PCA 1-gram plotly
    • PCA 2-gram & PCA 2-gram plotly
    • PCA 3-gram & PCA 3-gram plotly
  • Multidimensional Scaling 1000 MFW/MFF:
    • MDS 1-gram & MDS 1-gram plotly
    • MDS 2-gram & MDS 2-gram plotly
    • MDS 3-gram & MDS 3-gram plotly
  • Hierarchical Clusters 1000 MFW/MFF:
    • HCA 1-gram
    • HCA 2-gram
    • HCA 3-gram

Varia

• HJB - Showcase for a static and easy implementation of a possible digital edition of Hjertebogen. The implementation is based on R-markdown and uses the R-package flexdashboard for the sake of layout and style.

• Sovfilms Network - A Plotly-Plot of a Network for “Who plays with whom in Soviet films?” (“Wer spielt mit wem in sowjetischen Filmen?”). A university project. The files are available on my Github profile. The network layout is Fruchterman-Reingold (with basic paramaters, as in nx.spring_layout(G)).

• Data Science Day at U. Vienna and ACDH-CH Research Day 6 - Poster and supplementary materials for the event.

• My Rosalind profile: NKCZ at Rosalind.