Articles

We present three approaches for unsupervised grammar induction that are sensitive to data complexity and apply them to Klein and Man- ning’s Dependency Model with Valence. The first, Baby Steps, bootstraps itself via iterated learning of increasingly longer sentences and requires no initialization. This method sub- stantially exceeds Klein and …...

We present a novel deterministic dependency pars- ing algorithm that attempts to create the easiest arcs in the dependency structure first in a non-directional manner. Traditional deterministic parsing algorithms are based on a shift-reduce framework: they traverse the sentence from left-to-right and, at each step, per- form one of a …...

We present a method for disambiguating syn- tactic subjects from syntactic objects (a fre- quent ambiguity) in German sentences taken from an English-German bitext. We exploit the fact that subject and object are usually eas- ily determined in English. We show that a simple method disambiguates some subject- object ambiguities …...

We describe a method of incorporating task- specific cost functions into standard condi- tional log-likelihood (CLL) training of linear structured prediction models. Recently intro- duced in the speech recognition community, we describe the method generally for struc- tured models, highlight connections to CLL and max-margin learning for structured pre- diction …...

Most learning algorithms for undirected graphical models require complete inference over at least one instance before parameter up- dates can be made. SampleRank is a rank- based learning framework that alleviates this problem by updating the parameters during in- ference. Most semi-supervised learning algo- rithms also perform full inference on …...

We suggest improvements to a previously pro- posed framework for integrating Conditional Random Fields and Hidden Markov Models, dubbed a Crandem system (2009). The pre- vious authors’ work suggested that local la- bel posteriors derived from the CRF were too low-entropy for use in word-level automatic speech recognition. As an …...

We present Quantized Contour Modeling (QCM), a Bayesian approach to the classification of acoustic contours. We evaluate the performance of this tech- nique in the classification of prosodic events. We find that, on BURNC, this technique can success- fully classify pitch accents with 63.99% accuracy (.4481 CER), and phrase ending …...

This paper addresses the problem of learning phrase patterns for unsupervised speaker role classification. Phrase patterns are automati- cally extracted from large corpora, and redun- dant patterns are removed via a graph prun- ing algorithm. In experiments on English and Mandarin talk shows, the use of phrase pat- terns results …...

We address a key problem in grapheme-to- phoneme conversion: the ambiguity in map- ping grapheme units to phonemes. Rather than using single letters and phonemes as units, we propose learning chunks, or subwords, to re- duce ambiguity. This can be interpreted as learning a lexicon of subwords that has min- …...

We investigate in this paper the adequate unit of analysis for Arabic Mention Detection. We experiment different segmentation schemes with various feature-sets. Results show that when limited resources are available, models built on morphologically segmented data out- perform other models by up to 4F points. On the other hand, when …...

In this paper, we compare two novel methods for part of speech tagging of Arabic without the use of gold standard word segmentation but with the full POS tagset of the Penn Ara- bic Treebank. The first approach uses com- plex tags without any word segmentation, the second approach is …...

In this work, we try a hybrid methodology for language modeling where both morphological decomposition and factored language model- ing (FLM) are exploited to deal with the com- plex morphology of Arabic language. At the end, we are able to obtain from 3.5% to 7.0% relative reduction in word error …...

Phonetic string transduction problems, such as letter-to-phoneme conversion and name transliteration, have recently received much attention in the NLP community. In the past few years, two methods have come to dom- inate as solutions to supervised string trans- duction: generative joint n-gram models, and discriminative sequence models. Both ap- proaches …...

The task of identifying the language of text or utterances has a number of applications in natural language processing. Language iden- tification has traditionally been approached with character-level language models. How- ever, the language model approach crucially depends on the length of the text in ques- tion. In this paper, …...

This paper introduces a system designed for automatically generating personalized annota- tion tags to label Twitter user’s interests and concerns. We applied TFIDF ranking and TextRank to extract keywords from Twitter messages to tag the user. The user tagging pre- cision we obtained is comparable to the preci- sion of …...

In this paper, we focus on a recent Web trend called microblogging, and in particular a site called Twitter. The content of such a site is an extraordinarily large number of small textual messages, posted by millions of users, at ran- dom or in response to perceived events or sit- …...

We show how the Barzilay and Lapata entity- based coherence algorithm (2008) can be applied to a new, noisy data domain – student essays. We demonstrate that by combining Barzilay and Lapata’s entity-based features with novel features related to grammar errors and word usage, one can greatly improve the performance …...

Conversational Agents have been shown to be effective tutors in a wide range of educational domains. However, these agents are often ig- nored and abused in collaborative learning scenarios involving multiple students. In our work presented here, we design and evaluate interaction strategies motivated from prior re- search in small …...

In this paper we investigate how to identify initiation-response pairs in asynchronous, multi-threaded, multi-party conversations. We formulate the task of identifying initia- tion-response pairs as a pairwise ranking problem. A novel variant of Latent Semantic Analysis (LSA) is proposed to overcome a li- mitation of standard LSA models, namely that …...

In a corpus of expert tutoring dialogue, con- versation that is considered to be “off topic” (non-pedagogical) according to a previous coding scheme is explored for its value in tu- toring dynamics. Using the Linguistic Inquiry and Word Count (LIWC) tool, phases of tutor- ing categorized as “off topic” were …...