We introduce a method for learning to predict the following grammar and text of the ongoing translation given a source text. In our approach, predictions are offered aimed at reducing users’ burden on lexical and grammar choices, and improving productivity. The method involves learning syntactic phraseology and translation equivalents. At run-time, the source and its translation prefix are slic...

Most statistical machine translation (SMT) systems use phrase-to-phrase translations to capture local context information, leading to better lexical choices and more reliable word reordering. Long phrases capture more contexts than short phrases and result in better translation qualities. On the other hand, the increasing amount of bilingual data poses serious problems for storing all possible ...

This paper describes our participation to the Entrance Exams Task of CLEF 2015’s Question Answering Track. The goal is to answer multiple-choice questions on short texts. Our system first retrieves passages relevant to the question, through lexical expansion involving WordNet and word vectors. Then a tree edit model is used on graph representations of the passages and answer choices to extract ...

In this paper, we introduce a system, Sentence Planning Using Description, which generates collocations within the paradigm of sentence planning. SPUD simultaneously constructs the semantics and syntax of a sentence using a Lexicalized Tree Adjoining Grammar (LTAG). This approach captures naturally and elegantly the interaction between pragmatic and syntactic constraints on descriptions in a se...

Historical linguists have always known that some linguistic changes result from deliberate, conscious actions by speakers. But the general assumption has been that such changes are relatively trivial, confined mainly to the invention or borrowing of new words, changes in lexical semantics, and the adoption of a few structural features from a prestige dialect. The goal of this paper is to show t...

Speakers in dialogue tend to adapt to each other by starting to use similar lexical items, syntactic structures, or gestures. This behaviour, called alignment, may serve important cognitive, communicative and social functions (such as speech facilitation, grounding and rapport). Our aim is to enable and study the effects of these subtle aspects of communication in virtual conversational agents....

Sub-categorized arguments introduced by the French preposition pour has been under-studied in previous work, as can be seen from the incompleteness of existing lexical-syntactic resources in that regard. In this paper, we briefly introduce the various types of sub-categorization in pour, which are to be distinguished from occurrences of pour as a discourse connective. We describe how we added a...

HEAD DRIVEN STATISTICAL MODELS FOR NATURAL LANGUAGE PARSING Michael Collins Supervisor Professor Mitch Marcus Statistical models for parsing natural language have recently shown considerable suc cess in broad coverage domains Ambiguity often leads to an input sentence having many possible parse trees statistical approaches assign a probability to each tree thereby rank ing competing trees in or...

The extensive research done on the interrelationship between different social factors such as social class, gender and age and different linguistic variables has shown that these factors have important effects on the way language is used. Among these factors a special importance can be placed on age due to the role it plays in revealing different patterns of language use, such as the age-gradin...

The stories we tell about our lives unveil their content just as much as the lexical choices we make index a certain worldview, attitude, positionality, and relationship to reality. In essence, in narratives, individuals construct the self and denote personal identities. The available narrative identity studies have largely ignored the language employed by the bewitched while narrating their ex...