The most critical issue in generating and recognizing paraphrases is development of wide-coverage paraphrase knowledge. Previous work on paraphrase acquisition has collected lexicalized pairs of expressions; however, the results do not ensure full coverage of the various paraphrase phenomena. This paper focuses on productive paraphrases realized by general transformation patterns, and addresses...

We present a critique of language-based modelling for text input research, and propose an alternative inputbased approach. Current language-based statistical models are derived from large samples of text (corpora). However, this text reflects only the output, or final result, of the text input task. We argue that this weakens the utility of the model, because, (1) users’ language is typically q...

Assessing and understanding the impact of scattered and widespread events onto a mission is a pertinacious problem. Current approaches attempting to solve mission impact assessment employ score-based algorithms leading to spurious results. We identify a fourfold problem with score-based algorithms: (1) score-based algorithms enforce deep training of experts to employed frameworks for specificat...

Language models (LMs) are statistical models that calculate probabilities over sequences of words or other discrete symbols. Currently two major paradigms for language modeling exist: count-based n-gram models, which have advantages of scalability and test-time speed, and neural LMs, which often achieve superior modeling performance. We demonstrate how both varieties of models can be unified in...

The TNO system for multi-document summarisation is based on an extraction approach. For headline generation, we chose to extend our system to extract the most informative topical noun phrase. The cluster topic is defined as the most frequent term occurring in the most salient document sentences. The core of our system is a probabilistic model, which estimates the log-odds of salience based on a...

We introduce a statistical model for dialogues. We describe a dynamic programming algorithm that can be used to bracket a dialogue into segments and label each segment with its speech act. We evaluate the performance of the model. We also use this model for language modelling and get perplexity reduction.

The similarity language model is a statistical model that makes efficient use of long distance information when possible and falls back to standard ngram language model when not. To estimate the probability distribution of a given target context, each training example of the ngram model is retrieved and its similarity to the target context is estimated. In this work, this is done by performing ...

For statistical language model training, target task matched corpora are required. However, training corpora sometimes include both target task matched and unmatched sentences. In such a case, training set selection is effective for both model size reduction and model performance improvement. In this paper, training set selection method for statistical language model training is described. The ...

Statistical machine translation systems use a combination of one or more translation models and a language model. While there is a significant body of research addressing the improvement of translation models, the problem of optimizing language models for a specific translation task has not received much attention. Typically, standard word trigram models are used as an out-of-the-box component ...

The PANGLOSS Mark III system [Frederking et al. 94] was from the outset designed to be a symbolic, human-aided machine translation (MT) system. The need arose to rapidly adapt it for use as a fully-automated MT system. Our solution to this problem was to add a statistical English language model (ELM) to replace the most significant user activity, selecting between alternate translations produce...