Logo Lanfrica

RichardLitt/lrl

Domain:

natural language processing

Record type:

software
Creator:
Ric
Host:
For work concerning low resource languages. lrl === For script work concerning low resource languages. This does not include visualisations, Semantic Web data, or other random scripts that can be found in my other repositories. facebook-scraper -------- The Facebook scripts in here are for non-automatically harvesting data from Facebook groups, using manual AJAX querying and saving the source from the browser. It is not an automatic data collection scheme, nor a scraper, which makes it legal (afaik). A paper based is currently in progress. maltese-dict ------- I have developed a GUI and terminal-side dictionary program based on word lists I have access to; one from the internet, and a cleaned-up copy available via METASHARE on a CC BY-NC-SA license. I will presumably keep working on this throughout my time in Malta. maltese-* ------- In development. These are for courses at the University of Malta. One is a stemmer, based on the NLTK stemmers (Snowball, ISRI). One is a broken plural noun morphological analyser, based on previous work by Farrugia. The other is a chunker and basic code switching identifier, based on the work done in facebook-scraper and on Fabri's theoretical research on Maltese compounds.