I am working on context-base English to Marathi Machine Translation. I wanted to disambiguate nouns using Rule-Based MT approach for English to Marathi MT.
Thank you very much Quinn for responding and asking for claification.
The answers to both questions will be helpful to me but first I need to know how to select one meaning from multiple meaning considering its context in the sentence
There are many approaches to this class of problem. First of all is using close and further context in combination with the morphology of the sentence. Most of the sentences can be transformed into trees (using grammar) with leaves having sentence function appointed. Having this structure one can try to disambiguate the type of word. The other point is to build an ontology that would have "word-sense" to "term" relation defined and use it as input to your function.
Hi Goraksh, I think this link will be quite useful for you:
http://globalwordnet.org/wordnets-in-the-world/
The IndoWordNet and MarathiWordNet seem to have Marathi as a supported language.
As the others have said, to disambiguate words, it is important to build an ontology of every word with its word senses. Luckily for us, there are research groups who have done this for most languages, so all you have to do is use these resources for your purpose.
- Badges
- Science topic
- Similar topics
- Publishing