Class-based n-gram models of natural language
Computational Linguistics
Computational models for neuroscience
Unsupervised learning of the morpho-semantic relationship in MEDLINE®
BioNLP '07 Proceedings of the Workshop on BioNLP 2007: Biological, Translational, and Clinical Language Processing
Hi-index | 0.00 |
In natural language, the meaning of a lexeme often varies due to the specific surrounding context. Computational approaches to natural language processing can benefit from a reliable, long-range-context-dependent representation of the meaning of each lexeme that appears in a given sentence. We have developed a general new technique that produces a context-dependent 'meaning' representation for a lexeme in a specific surrounding context. The 'meaning' of a lexeme in a specific context is represented by a list of semantically replaceable elements the members of which are other lexemes from our experimental lexicon. We have performed experiments with a lexicon composed of individual English words and also with a lexicon of individual words and selected phrases. The resulting lists can be used to compare the 'meaning' of conceptual units (individual words or frequently-occurring phrases) in different contexts and also can serve as features for machine learning approaches to classify semantic roles and relationships.