Articles

Letter-to-phoneme conversion generally requires aligned training data of letters and phonemes. Typically, the align- ments are limited to one-to-one align- ments. We present a novel technique of training with many-to-many alignments. A letter chunking bigram prediction man- ages double letters and double phonemes automatically as opposed to preprocess- ing with …...

Scientific papers revolve around cita- tions, and for many discourse level tasks one needs to know whose work is being talked about at any point in the discourse. In this paper, we in- troduce the scientific attribution task, which links different linguistic expres- sions to citations. We discuss the suitability …...

Sentiment analysis seeks to characterize opinionated or evaluative aspects of nat- ural language text. We suggest here that appraisal expression extraction should be viewed as a fundamental task in sentiment analysis. An appraisal expression is a tex- tual unit expressing an evaluative stance towards some target. The task is to …...

We address the problem of analyzing mul- tiple related opinions in a text. For in- stance, in a restaurant review such opin- ions may include food, ambience and ser- vice. We formulate this task as a multiple aspect ranking problem, where the goal is to produce a set of numerical …...

We propose a method for extracting se- mantic orientations of phrases (pairs of an adjective and a noun): positive, negative, or neutral. Given an adjective, the seman- tic orientation classification of phrases can be reduced to the classification of words. We construct a lexical network by con- necting similar/related words. …...

Motivated by psycholinguistic findings, we are currently investigating the role of eye gaze in spoken language understand- ing for multimodal conversational sys- tems. Our assumption is that, during hu- man machine conversation, a user’s eye gaze on the graphical display indicates salient entities on which the user’s atten- tion is …...

Past approaches for using reinforcement learning to derive dialog control policies have assumed that there was enough col- lected data to derive a reliable policy. In this paper we present a methodology for numerically constructing confidence inter- vals for the expected cumulative reward for a learned policy. These intervals are …...

This paper studies methods that automat- ically detect action-items in e-mail, an important category for assisting users in identifying new tasks, tracking ongoing ones, and searching for completed ones. Since action-items consist of a short span of text, classifiers that detect action-items can be built from a document-level or a …...

Reinforcement learning gives a way to learn under what circumstances to per- form which actions. However, this ap- proach lacks a formal framework for spec- ifying hand-crafted restrictions, for speci- fying the effects of the system actions, or for specifying the user simulation. The in- formation state approach, in contrast, …...

We propose a new framework for supervised ma- chine learning. Our goal is to learn from smaller amounts of supervised training data, by collecting a richer kind of training data: annotations with “ra- tionales.” When annotating an example, the hu- man teacher will also highlight evidence support- ing this annotation—thereby …...

In unsupervised learning, where no train- ing takes place, one simply hopes that the unsupervised learner will work well on any unlabeled test collection. How- ever, when the variability in the data is large, such hope may be unrealistic; a tuning of the unsupervised algorithm may then be necessary in …...

We describe Castanet, an algorithm for auto- matically generating hierarchical faceted meta- data from textual descriptions of items, to be in- corporated into browsing and navigation inter- faces for large information collections. From an existing lexical database (such as WordNet), Castanet carves out a structure that reflects the contents of …...

Standard pairwise coreference resolution systems are subject to errors resulting from their performing anaphora identifi- cation as an implicit part of coreference resolution. In this paper, we propose an integer linear programming (ILP) for- mulation for coreference resolution which models anaphoricity and coreference as a joint task, such that each …...

Currently there are several approaches to machine translation (MT) based on differ- ent paradigms; e.g., phrasal, hierarchical and syntax-based. These three approaches yield similar translation accuracy despite using fairly different levels of linguistic knowledge. The availability of such a variety of systems has led to a growing interest toward finding …...

Information retrieval systems are fre- quently required to handle long queries. Simply using all terms in the query or re- lying on the underlying retrieval model to appropriately weight terms often leads to ineffective retrieval. We show that re- writing the query to a version that com- prises a small …...

This paper introduces a novel evaluation framework for question series and em- ploys it to explore the effectiveness of QA and IR systems at addressing users’ infor- mation needs. The framework is based on the notion of recall curves, which char- acterize the amount of relevant informa- tion contained within …...

Graph-based semi-supervised learning has recently emerged as a promising approach to data-sparse learning problems in natu- ral language processing. All graph-based algorithms rely on a graph that jointly rep- resents labeled and unlabeled data points. The problem of how to best construct this graph remains largely unsolved. In this paper …...

This paper describes a method for generat- ing sense-tagged data using Wikipedia as a source of sense annotations. Through word sense disambiguation experiments, we show that the Wikipedia-based sense annotations are reliable and can be used to construct accurate sense classifiers. 1...

We present a sentence compression system based on synchronous context-free grammars (SCFG), following the successful noisy-channel approach of (Knight and Marcu, 2000). We define a head- driven Markovization formulation of SCFG dele- tion rules, which allows us to lexicalize probabili- ties of constituent deletions. We also use a robust approach …...

This paper explores the use of statisti- cal machine translation (SMT) methods for tactical natural language generation. We present results on using phrase-based SMT for learning to map meaning repre- sentations to natural language. Improved results are obtained by inverting a seman- tic parser that uses SMT methods to map …...