Task ideas for Google Code-in/Add words to monolingual dictionary

From Apertium
< Task ideas for Google Code-in
Revision as of 00:25, 9 December 2019 by Firespeaker (talk | contribs) (Created page with "# '''Select a language module''', ideally such that the language is a language you know. # '''Install Apertium''' locally from nightlies ([[Installation#Installing:_a_summary|...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search
  1. Select a language module, ideally such that the language is a language you know.
  2. Install Apertium locally from nightlies (instructions here); clone the relevant language module from GitHub; compile it; and check that it works. Alternatively, get Apertium VirtualBox and update, check out & compile the language pair.
  3. Using a large enough corpus of representative text in the language (e.g. plain text taken from Wikipedia, newspapers, literature, etc.) detect the 200 most frequent unknown words (words in the source document which are not in the dictionary). See below for information about how to do this. Note: the beginner version of this task only requires 50 words.
  4. Add the words to the monolingual dictionary (the appropriate .dix or .lexc file) so that they are not unknown anymore. Make sure to categorise stems correctly (this can be hard, so please check with your mentor if you're unsure about anything!).
  5. Compile and test again
  6. Submit a pull request to the GitHub repository with your updates.

How to find the most frequent unknowns

Paraphrased from Unhammer:

  • analyse your corpus, make it one word per line, grab only the ones with * at the start, sort, count number of hits per word, sort again
  • e.g. zcat corpus.txt.gz | apertium -d . ron-morph | tr ' ' '\n' | grep '^\*' | sort |uniq -d |sort -n >hitlist
<Unhammer> hitlist will be unknowns sorted by frequency, but you might have to
           skip a couple that are "strange" or difficult to add
<Unhammer> and that's ok, as long as you start from the most frequent and work
           your way down