Unsupervised Graph-based Word Sense Disambiguation Using Lexical Relation of WordNet
Word Sense Disambiguation (WSD) is one of tasks in the Natural Language Processing that uses to identifying the sense of words in context. To select the correct sense, we can use many approach. This paper uses a tree and graph-connectivity structure for finding the correct senses. This algorithm has few parameters and does not require sense-annotated data for training. Performance evaluation on standard datasets showed it has the better accuracy than many previous graph base algorithms and decreases elapsed time.
Keywords: word sense disambiguation, tree, Graph connectivity.
Download Full-Text
ABOUT THE AUTHORS
Ehsan Hessami
Islamic Azad University, Qazvin Branch
Faribourz Mahmoudi
Islamic Azad University, Qazvin Branch
Islamic Azad Amir Hossien JadidinejadUniversity, Qazvin Branch
Ehsan Hessami
Islamic Azad University, Qazvin Branch
Faribourz Mahmoudi
Islamic Azad University, Qazvin Branch
Islamic Azad Amir Hossien JadidinejadUniversity, Qazvin Branch
