A simple word sense disambiguation library written in scala with a java api
Wordnet 3.0 (using extJWNL)
StanforNLP
Neo4j graph database and scala-gremlin
On input text, perform part-of-speech tagging
On tagged text, derive intersection points for the different possible synsets of the word
Disambiguate words that can be associated with one or more intersection points
For the remaining words, disambiguate to most popular context for that part of speech
Add support for word compoundification
Provide a simple play webapp to deploy this as a REST service
Research into other alternatives to using StanfordNLP
Research into using other graph databases or in-memory graphs like TinkerGraph etc.