Word reordering is one of the fundamental problems of machine translation, and an important factor of its quality and efficiency. In this paper, we introduce a novel reordering model based on an innovative structure, named, phrasal dependency tree including syntactical and statistical information in context of a log-linear model. The phrasal dependency tree is a new modern syntactic structure based on dependency relations between contiguous non-syntactic phrases. In comparison with well-known and popular reordering models such as the distortion, lexicalized and hierarchical models, the experimental study demonstrates the superiority of our model regarding to the different evaluation measures. We evaluated the proposed model on a PersianEnglish SMT system. On average our model retrieved a significant impact on precision with comparable recall value respect to the lexicalized and distortion models, and is found to be effective for medium and long-distance reordering.

This paper proposes ELEXR, a novel metric to evaluate machine translation (MT). In our proposed method, we extract lexical co-occurrence relationships of a given reference translation (Ref) and its corresponding hypothesis sentence using hyperspace analogue to language space matrix. Then, for each term appearing in these two sentences, we convert the co-occurrence information into a conditional probability distribution. Finally, by comparing the conditional probability distributions of the words held in common by Ref and the candidate sentence (Cand) using Kullback-Leibler divergence, we can score the hypothesis. ELEXR can evaluate MT by using only one Ref assigned to each Cand without incorporating any semantic annotated resources like WordNet. Our experiments on eight language pairs of WMT 2011 submissions show that ELEXR outperforms baselines, TER and BLEU, on average at system-level correlation with human judgments. It achieves average Spearman’s rho correlation of about 0.78, Kendall’s tau correlation of about 0.66 and Pearson’s correlation of about 0.84, corresponding to improvements of about 0.04, 0.07 and 0.06 respectively over BLEU, the best baseline.

This paper presents a method for converting back and forth between the Perso-Arabic and a romanized writing system for Persian. Given a word in one writing system, we use finite state transducers to generate morphological analysis for the word that is subsequently used to regenerate the orthography of the word in the other writing system. The system has been implemented in XFST and LEXC.

This paper presents a method for converting back and forth between the Perso-Arabic and a romanized writing system for Persian. Given a word in one writing system, we use finite state transducers to generate morphological analysis for the word that is subsequently used to regenerate the orthography of the word in the other writing system. The system has been implemented in XFST and LEXC.

This paper presents, Dabire, a romanized transcription scheme which is based on the phonology and morphophonology of Persian. Dabire uses an extended Latin alphabet and a number of conventions with the aim of providing a simple, consistent and easy to learn writing system.

This paper describes a syllabification based conversion method for converting romanized Persian text to the traditional Arabic-based writing system. The system is implemented in Xerox XFST and relies on rule based conversion of words rather than using morphological analysis. The paper presents a brief evaluation of the accuracy of the transcriptions generated by the method.

This paper proposes the use of description logics as a representational framework for learning composite concepts. Description logics are restricted variants of first-order logic providing a form of logical bias that dates back to semantic networks. Some recent work investigates concept learning in the context of these formalisms. Also, having recognized the importance of part-whole hierarchies in commonsense reasoning, researchers have started to incorporate part-of reasoning into description logics. In our approach we represent composite concepts in such a formalism. On one hand we have a relatively rich representation language with an infinite space of possible concepts. On the other hand we have special constructs for handling part-of relations that can be used in the learning algorithm to reduce the overall search space.

Problem solving involves search. In AI we try to find ways of avoiding or minimizing search. An effective approach is to exploit knowledge of the problem domain. Such knowledge often takes the form of a set of constraints. In general, a constraint represents a required relationship among some variables. For this reason we usually assume the existence of some machinery that can enforce a given set of constraints and resolve the conflicts that arise when these are violated. Programming systems based on constraints have been successfully applied to circuit analysis, design and simulation, scene analysis, plan generation, hardware diagnosis and qualitative reasoning. This thesis presents ICONStraint, a programming language based on the constraints paradigm of computation and gives a characterization of consistency conditions and the operations for ensuring consistency in constraint networks designed in the language. In ICONStraint we represent constraint systems in dependency structures and use reason maintenance, local propagation and dependency directed backtracking for computation and consistency maintenance. ICONStraint has been implemented in Interlisp-D.