shamotskyi's picture
Create README.md
dceed35 verified
|
raw
history blame
1.38 kB
metadata
language:
  - en
license: cc-by-nc-4.0
tags:
  - news

For each article, its text and titles are given, as well as masked text and title (with all digits replaced with "X"). The indexes of the similar articles refer to the ids in the ukr_pravda_2y dataset (TODO check if this is factually correct).

Then, as ML eval task, a choice of 10 masked titles from similar articles are given (including the 'real' one). The label column points to the index of the correct masked title.

Similarity of articles is a dead-simple cosine distance over binary vectors of the articles tags:

  • a vector is built using spacy CountVectorizer, with 0 if the tag is absent and 1 if present
  • similarity is cosine distance between these vectors of two articles
  • the 10 most similar articles' titles are taken

NB this simple similarity may be suboptimal, because there are MANY UP articles with the exact same tags (ergo all with similarity 1.0 to the source), and there may be more similar articles (by human intuition as well as by any more reasonable metric) than the ones actually chosen.

This is done in the context of my Master's thesis, better documentation will follow soon.