We address the problem of unknown words, also known as out of vocabulary (OOV) words, in machine translation of low resource languages. Our technique comprises a combination of methods, inspired by the common OOV types observed. We also design evaluation techniques for measuring coverage of OOVs achieved and integrate the new translation candidates in a Statistical Machine Translation (SMT) sys...

This paper studies the sampling strategies for the Expert Network (EexNet), a statistical learning system used for patient record classification at the Mayo Clinic. The goal is to achieve high accuracy classification at an affordable computational cost in very large applications. The learning curves of ExpNet were observed with respect to the choice of training resources, the size, vocabulary c...

We propose a methodology that adapts graph embedding techniques (DeepWalk (Perozzi et al., 2014) and node2vec (Grover and Leskovec, 2016)) as well as crosslingual vector space mapping approaches (Least Squares and Canonical Correlation Analysis) in order to merge the corpus and ontological sources of lexical knowledge. We also perform comparative analysis of the used algorithms in order to iden...

WoNeF, an improved, extended and evaluated automatic French translation of WordNet Identifying the various possible meanings of each word of the vocabulary is a difficult problem that requires a lot of manual work. It has been tackled by the WordNet lexical semantics database in English, but there are still few resources available for other languages. Automatic translations of WordNet have been...

This paper studies training set sampling strategies in the context of statistical learning for text cate-gorization. It is argued sampling strategies favoring common categories is superior to uniform coverage or mistake-driven approaches, if performance is measured by globally assessed precision and recall. The hypothesis is empirically validated by examining the performance of a nearest neighb...

this study was designed to find which one of the three different presentations, i.e. input, input-output, and output-input, will be more effective in iranian efl learners' vocabulary acquisitions. to this end, first 54 out of 64 female students, aged from 19 to 23 years, with an average of 21, were selected out of starter-level efl learners at the university of tarbiat moalem in bandar abb...

This paper studies training set sampling strategies in the context of statistical learning for text categorization. It is argued sampling strategies favoring common categories is superior to uniform coverage or mistake-driven approaches, if performance is measured by globally assessed precision and recall. The hypothesis is empirically validated by examining the performance of a nearest neighbo...

The Unified Medical Language System (UMLS) was examined to determine its coverage of clinical laboratory terminology in use at the Columbia-Presbyterian Medical Center (CPMC). The Metathesaurus (Meta-1) contains exact matches for 30% of 1460 CPMC laboratory terms and near matches for an additional 42%, with better coverage of atomic-level concepts ("substance" terms) than complex ones (tests an...