Feel free to edit/comment/spam/anything here
IMHO the LS problem can be reduced to a classification problem:
the context could be a text frame, a bag of words, a tfidf-labelled array etc.
the classification problem can be solved in various ways: support vector machines, naive-bayes classifier, decision tree etc.
Some Bookmarks (please feel free to add more)
Using UMLS Concept Unique Identifiers (CUIs) for Word Sense Disambiguation in the Biomedical Domain: http://www.d.umn.edu/~tpederse/Pubs/amia07.pdf
Word Sense Disambiguation - Algorithms and Applications: http://www.wsdbook.org/
Word Sense Disambiguation: The State of the Art: http://sites.univ-provence.fr/~veronis/pdf/1998wsd.pdf
Word Sense Disambiguation (slide from the "Linguaggi e Traduttori" class): http://www.di.uniba.it/~semeraro/LT/WSD.pdf
Perl scripts doing WSD and mapping on UMLS ontologies: http://cuitools.sourceforge.net/
Nice ACM survey on WSD: http://www.dsi.uniroma1.it/~navigli/pubs/ACM_Survey_2009_Navigli.pdf
Verb Semantics and Lexical Selection: http://www.ldc.upenn.edu/acl/P/P94/P94-1019.pdf