The IAMTC project (Interlingual Annotation of Multilingual Translation Corpora) is developing an interlingual representation framework for annotation of parallel corpora (English paired with Arabic, French, Hindi, Japanese, Korean, and Spanish) with deep-semantic representations. In particular, we are investigating meaning equivalent paraphrases involving conversives and non-literal language us...

Many corpus-based machine translation systems require parallel corpora. In this paper, we present a word-for-word glossing algorithm that requires only a source language corpus. To gloss a word, we first identify its similar words that occurred in the same context in a large corpus. We then determine the gloss by maximizing the similarity between the set of contextually similar words and the di...

Dissertation title : Discriminative Interlingual Representations Jagadeesh Jagarlamudi, Doctor of Philosophy, 2013 Dissertation advised by: Hal Daumé III Department of Computer Science The language barrier in many multilingual natural language processing (NLP) tasks can be overcome bymapping objects from different languages (“views”) into a common low-dimensional subspace. For example, the name...

Abstract In today’s multilingual lexical databases, the majority of world’s languages are under-represented. Beyond a mere issue resource incompleteness, we show that existing databases have structural limitations result in reduced expressivity on culturally-specific words and mapping them across languages. particular, meaning space dominant languages, such as English, is represented more accur...

The popularity of information retrieval has led users to a new problem: nding which text databases (out of thousands of candidate choices) are the most relevant to a user. Answering a given query with a list of relevant databases is the text database discovery problem. The rst part of this paper presents a practical method for attacking this problem based on estimating the result size of a quer...

We propose a novel subband-based S-curve transformation for increasing the perceived contrast of images, and use it to explore the relation between perceived gloss and perceived contrast of natural textures. The proposed transformation makes minimal assumptions on lighting conditions and does not require prior knowledge of surface geometry. Through a series of subjective experiments with both c...

BACKGROUND The purpose of this study was to evaluate orthodontic debonding methods by comparing the surface roughness and enamel morphology of teeth after applying two different debonding methods and three different polishing techniques. METHODS Forty eight human maxillary premolars, extracted for orthodontic reasons, were randomly divided into three groups. Brackets were bonded to teeth with...

Word Sense Disambiguation (WSD) aims to predict the correct sense of a word given its context. This problem is extreme importance in Arabic, as written words can be highly ambiguous; 43% diacritized have multiple interpretations and percentage increases 72% for non-diacritized words. Nevertheless, most Arabic text does not diacritical marks. Gloss-based WSD methods measure semantic similarity o...

For the rst interactive Cross-Language Evaluation Forum, the Maryland team focused on comparison of term-for-term gloss translation with full machine translation for the document selection task. The results show that (1) searchers are able to make relevance judgments with translations from either approach, and (2) the machine translation system achieved better eeectiveness than the gloss transl...