This paper presents a cluster-based text categorization system which uses class distributional clustering of words. We propose a new clustering model which considers the global in...
Computing the degree of semantic relatedness of words is a key functionality of many language applications such as search, clustering, and disambiguation. Previous approaches to c...
Kira Radinsky, Eugene Agichtein, Evgeniy Gabrilovi...
Standard cursive handwriting recognition is based on a language model, mostly a lexicon of possible word hypotheses or character n-grams. The result is a list of word alternatives...
In off-line handwriting recognition, classifiers based on hidden Markov models (HMMs) have become very popular. However, while there exist well-established training algorithms, s...
In this paper, we introduce an on-line Arabic handwritten recognition system based on new stroke segmentation algorithm. The proposed algorithm uses an over segmentation method th...