User:Frankier

Basic info

Full name: Frankie Robertson

Email: $first@$second.name

Web: http://frankie.robertson.name

Phone: +358 46576679

IRC: frankier

About me/blurb (cribbed from GSOC application)

My interest in Natural Language Processing was initially sparked by my attempts to learn the Finnish language which got me pondering language and language learning in general quite a bit. My long term vision that drives my interest is that I think eventually we can apply some of these tools to improve the experience of language learning. The idea is something along the lines of “if we can make it a model for a computer - we can get the computer to teach it to a human”. I don’t think I’m alone in having this long term vision - for example the VISL project http://beta.visl.sdu.dk/ and the Oahpa! Project http://oahpa.no/ have worked in this direction (but I think there is a lot more to be done and different potential ways to apply NLP techniques to help language learners).

Since I have experience in software engineering, programming and computer science, NLP feels like a very direct and natural way to engage with linguistics and linguistic issues.

One thing that interests me about rule based approaches in comparison to statistical/ML approaches to machine translation and NLP is that a rule based system can explain itself to a language learner (some statistical/ML approaches can learn rules - such hybrid systems might also be able to explain themselves).

Some pages

• Code style - trying to document existing Apertium code style and make steps towards a consistent style for Apertium code
• Jenkins - CI for Apertium code and language data
• CG hybrid tagging
• Perceptron tagger