Datasets:

yxchar

/

citation_intent-tlm

like 2

We offer a theorem that highlights the broad applicability of these modeling techniques .4 If f ( input , output ) is a weighted regular relation , then the following statements are equivalent : ( 1 ) f is a joint probabilistic relation ; ( 2 ) f can be computed by a Markovian FST that halts with probability 1 ; ( 3 ) f can be expressed as a probabilistic regexp , i.e. , a regexp built up from atomic expressions a : b ( for a E E U -LCB- E -RCB- , b E A U -LCB- E -RCB- ) using concatenation , probabilistic union + p , and probabilistic closure * p. For defining conditional relations , a good regexp language is unknown to us , but they can be defined in several other ways : ( 1 ) via FSTs as in Fig. 1c , ( 2 ) by compilation of weighted rewrite rules ( Mohri and Sproat , 1996 ) , ( 3 ) by compilation of decision trees ( Sproat and Riley , 1996 ) , ( 4 ) as a relation that performs contextual left-to-right replacement of input substrings by a smaller conditional relation ( Gerdemann and van Noord , 1999 ) ,5 ( 5 ) by conditionalization of a joint relation as discussed below .

We use the Clark and Curran ( 2007 ) CCG parser to analyse the sentence before and after paraphrasing .

For example , consider a relational description ( cfXXX , Dale and Haddock 1991 ) involving a gradable adjective , as in the dog in the large shed .

Previous versions of our work , as described in Bachenko et al. ( 1986 ) also assume that phrasing is dependent on predicate-argument structure .

like information extraction ( Yates and Etzioni , 2009 ) and textual entailment ( Berant et al. , 2010 ) .

The inclusion of the coreference task in the Sixth and Seventh Message Understanding Conferences ( MUC-6 and MUC-7 ) gave a considerable impetus to the development of coreference resolution algorithms and systems , such as those described in Baldwin et al. ( 1995 ) , Gaizauskas and Humphreys ( 1996 ) , and Kameyama ( 1997 ) .

Lee et al. ( 2012 ) model entity coreference and event coreference jointly ; Durrett and Klein ( 2014 ) consider joint coreference and entity-linking .

However , learning-based resolvers have not been able to benefit from having an SC agreement feature , presumably because the method used to compute the SC of an NP is too simplistic : while the SC of a proper name is computed fairly accurately using a named entity ( NE ) recognizer , many resolvers simply assign to a common noun the first ( i.e. , most frequent ) WordNet sense as its SC ( e.g. , Soon et al. ( 2001 ) , Markert and Nissim ( 2005 ) ) .

This includes work on generalized expectation ( Mann and McCallum , 2010 ) , posterior regularization ( Ganchev et al. , 2010 ) and constraint driven learning ( Chang et al. , 2007 ; Chang et al. , 2010 ) .

The final machine is a trigram language model , specifically a Kneser-Ney ( Chen and Goodman , 1998 ) based backoff language model .

Hovy has described another text planner that builds similar plans ( Hovy 1988b ) .

In particular , since we treat each individual speech within a debate as a single `` document '' , we are considering a version of document-level sentiment-polarity classification , namely , automatically distinguishing between positive and negative documents ( Das and Chen , 2001 ; Pang et al. , 2002 ; Turney , 2002 ; Dave et al. , 2003 ) .

transition-based dependency parsing framework ( Nivre , 2008 ) using an arc-eager transition strategy and are trained using the perceptron algorithm as in Zhang and Clark ( 2008 ) with a beam size of 8 .

Just as easily , we can model link types that coincide with entries in an on-line bilingual dictionary separately from those that do not ( cfXXX Brown et al. 1993 ) .

As an alternative , we rely on PubMed to retrieve an initial set of hits that we then postprocess in greater detail -- this is the standard pipeline architecture commonly employed in other question-answering systems ( Voorhees and Tice 1999 ; Hirschman and Gaizauskas 2001 ) .

( Och and Ney , 2002 ; Blunsom et al. , 2008 ) used maximum likelihood estimation to learn weights for MT. ( Och , 2003 ; Moore and Quirk , 2008 ; Zhao and Chen , 2009 ; Galley and Quirk , 2011 ) employed an evaluation metric as a loss function and directly optimized it .

We also compare the results with the output generated by the statistical translation system GIZA + + / ISI ReWrite Decoder ( AlOnaizan et al. , 1999 ; Och and Ney , 2000 ; Germann et al. , 2001 ) , trained on the same parallel corpus .

Some efforts have tackled tasks such as automatic image caption generation ( Feng and Lapata , 2010a ; Ordonez et al. , 2011 ) , text illustration ( Joshi et al. , 2006 ) , or automatic location identification of Twitter users ( Eisenstein et al. , 2010 ; Wing and Baldridge , 2011 ; Roller et al. , 2012 ) .

Since we are not generating from the model , this does not introduce difficulties ( Klein and Manning , 2002 ) .

The extraction procedure consists of three steps : First , the bracketing of the trees in the Penn Treebank is corrected and extended based on the approaches of Magerman ( 1994 ) and Collins ( 1997 ) .

As they are required to enable test subjects to distinguish between senses , we use artificial glosses composed from synonyms and hypernyms as a surrogate , e.g. for brother : `` brother , male sibling '' vs. `` brother , comrade , friend '' ( Gurevych , 2005 ) .

Over the last decade there has been a lot of interest in developing tutorial dialogue systems that understand student explanations ( Jordan et al. , 2006 ; Graesser et al. , 1999 ; Aleven et al. , 2001 ; Buckley and Wolska , 2007 ; Nielsen et al. , 2008 ; VanLehn et al. , 2007 ) , because high percentages of selfexplanation and student contentful talk are known to be correlated with better learning in humanhuman tutoring ( Chi et al. , 1994 ; Litman et al. , 2009 ; Purandare and Litman , 2008 ; Steinhauser et al. , 2007 ) .

Dalrymple ( 2001 ) argues that there are cases , albeit exceptional ones , in which constraints on syntactic category are an issue in subcategorization .

Corpus frequency : ( Vosse , 1992 ) differentiates between misspellings and neologisms ( new words ) in terms of their frequency .

We use an in-house developed hierarchical phrase-based translation ( Chiang , 2005 ) as our baseline system , and we denote it as In-Hiero .

The types of sentences accepted are essentially those accepted by the original NLC grammar , imperative sentences with nested noun groups and conjunctions ( Ballard 1979 ) .

The task we used to compare different generalisation techniques is similar to that used by Pereira et al. ( 1993 ) and Rooth et al. ( 1999 ) .

One of the proposed methods to extract paraphrases relies on a pivot-based approach using phrase alignments in a bilingual parallel corpus ( Bannard and Callison-Burch , 2005 ) .

The Nash arbitration plan , for example , would allow a doubly graded description whenever the product of the Values for the referent r exceeds that of all distractors ( Nash 1950 ; cfXXX Gorniak and Roy 2003 ; Thorisson 1994 , for other plans ) .

Turney ( 2001 ) extracts word co-occurrence probabilities from unlabelled text collected from a web crawler .

Notable early papers on graph-based semisupervised learning include Blum and Chawla ( 2001 ) , Bansal et al. ( 2002 ) , Kondor and Lafferty ( 2002 ) , and Joachims ( 2003 ) .

Due to this inherent ambiguity , manual annotations usually distinguish between sure correspondences for unambiguous translations , and possible , for ambiguous translations ( Och and Ney 2003 ) .

Thus for instance , ( Copestake and Flickinger , 2000 ; Copestake et al. , 2001 ) describes a Head Driven Phrase Structure Grammar ( HPSG ) which supports the parallel construction of a phrase structure ( or derived ) tree and of a semantic representation and ( Dalrymple , 1999 ) show how to equip Lexical Functional grammar ( LFG ) with a glue semantics .

The contextual interpreter then uses a reference resolution approach similar to Byron ( 2002 ) , and an ontology mapping mechanism ( Dzikovska et al. , 2008a ) to produce a domain-specific semantic representation of the student 's output .

only the available five relative scopings of the quantifiers are produced ( Hobbs and Shieber 1987 , 47 ) , but without the need for a free variable constraint -- the HOU algorithm will not produce any solutions in which a previously bound variable becomes free ; • the equivalences are reversible , and thus the above sentences cart be generated from scoped logical forms ; • partial scopings are permitted ( see Reyle [ 19961 ) • scoping can be freely interleaved with other types of reference resolution ; • unscoped or partially scoped forms are available for inference or for generation at every stage .

In the transducers produced by the training method described in this paper , the source and target positions are in the set -LCB- -1 , 0,1 -RCB- , though we have also used handcoded transducers ( Alshawi and Xia 1997 ) and automatically trained transducers ( Alshawi and Douglas 2000 ) with a larger range of positions .

Similar approaches are being explored for parsing ( Steedman , Hwa , et al. 2003 ; Hwa et al. 2003 ) .

In particular , since we treat each individual speech within a debate as a single `` document '' , we are considering a version of document-level sentiment-polarity classification , namely , automatically distinguishing between positive and negative documents ( Das and Chen , 2001 ; Pang et al. , 2002 ; Turney , 2002 ; Dave et al. , 2003 ) .

would be chunked as follows ( Tjong Kim Sang and Buchholz , 2000 ) : [ NP He ] [ VP reckons ] [ NP the current account deficit ] [ VP will narrow ] [ PP

The flexible architecture we have presented enables interesting future research : ( i ) a straightforward improvement is the use of lexical similarity to reduce data sparseness , e.g. ( Basili et al. , 2005 ; Basili et al. , 2006 ; Bloehdorn et al. , 2006 ) .

5An alternative strategy to step ( 4 ) is to perform a database lookup based on the ambiguous query and summarize the results ( Litman et al. , 1998 ) , which we leave for future work .

Such approaches have been tried recently in restricted cases ( McCallum et al. , 2000 ; Eisner , 2001b ; Lafferty et al. , 2001 ) .

Typed feature grammars can be used as the basis for implementations of Head-driven Phrase Structure Grammar ( HPSG ; Pollard and Sag , 1994 ) as discussed in ( Gotz and Meurers , 1997a ) and ( Meurers and Minnen , 1997 ) .

It is therefore no surprise that early attempts at response automation were knowledge-driven ( Barr and Tessler 1995 ; Watson 1997 ; Delic and Lahaix 1998 ) .

A more subtle example is weighted FSAs that approximate PCFGs ( Nederhof , 2000 ; Mohri and Nederhof , 2001 ) , or to extend the idea , weighted FSTs that approximate joint or conditional synchronous PCFGs built for translation .

ones , DIRT ( Lin and Pantel , 2001 ) , VerbOcean ( Chklovski and Pantel , 2004 ) , FrameNet ( Baker et al. , 1998 ) , and Wikipedia ( Mehdad et al. , 2010 ; Kouylekov et al. , 2009 ) .

The Ruby on Rails ( 2006 ) framework permits us to quickly develop web applications without rewriting common functions and classes .

We further add rules for combining with punctuation to the left and right and allow for the merge rule X → X X of Clark and Curran ( 2007 ) .

Politically-oriented text Sentiment analysis has specifically been proposed as a key enabling technology in eRulemaking , allowing the automatic analysis of the opinions that people submit ( Shulman et al. , 2005 ; Cardie et al. , 2006 ; Kwon et al. , 2006 ) .

Moschitti et al. ( 2005 ) has made some preliminary attempt on the idea of hierarchical semantic

Morris and Hirst ( 2004 ) pointed out that many relations between words in a text are non-classical ( i.e. other than typical taxonomic relations like synonymy or hypernymy ) and therefore not covered by semantic similarity .

Some of the intuitions we associate with this notion have been very well expressed by Turner ( 1987 , pp. 7-8 ) : ... Semantics is constrained by our models of ourselves and our worlds .

In 2009 , the second WePS campaign showed similar trends regarding the use of NE features ( Artiles et al. , 2009 ) .

The head words can be automatically extracted using a heuristic table lookup in the manner described by Magerman ( 1994 ) .

Reiter describes a pipelined modular approach as a consensus architecture underlying most recent work in generation ( Reiter 1994 ) .

For example , the suite of LT tools ( Mikheev et al. , 1999 ; Grover et al. , 2000 ) perform tokenization , tagging and chunking on XML marked-up text directly .

Liu et al. ( 2005 ) , Meral et al. ( 2007 ) , Murphy ( 2001 ) , Murphy and Vogel ( 2007 ) and Topkara et al. ( 2006a ) all belong to the syntactic transformation category .

Brockmann and Lapata ( 2003 ) have showed that WordNet-based approaches do not always outperform simple frequency-based models , and a number of techniques have been recently proposed which may offer ideas for refining our current unsupervised approach ( Erk , 2007 ; Bergsma et al. , 2008 ) .

The RenTAL system is implemented in LiLFeS ( Makino et al. , 1998 ) 2 .

In the future , we hope to evaluate the automatic annotations and extracted lexicon against Propbank ( Kingsbury and Palmer 2002 ) .

Each set of translations is stored separately , and for each set the `` marker hypothesis '' ( Green 1979 ) is used to segment the phrasal lexicon into a `` marker lexicon . ''

One way to increase the precision of the mapping process is to impose some linguistic constraints on the sequences such as simple noun-phrase contraints ( Gaussier , 1995 ; Kupiec , 1993 ; hua Chen and Chen , 94 ; Fung , 1995 ; Evans and Zhai , 1996 ) .

Although this study falls under the general topic of discourse modeling , our work differs from previous attempts to characterize text in terms of domainindependent rhetorical elements ( McKeown , 1985 ; Marcu and Echihabi , 2002 ) .

`` Coherence , '' as outlined above , can be understood as a declarative ( or static ) version of marker passing ( Hirst 1987 ; Charniak 1983 ) , with one difference : the activation spreads to theories that share a predicate , not through the IS-A hierarchy , and is limited to elementary facts about predicates appearing in the text .

Task properties Determining whether or not a speaker supports a proposal falls within the realm of sentiment analysis , an extremely active research area devoted to the computational treatment of subjective or opinion-oriented language ( early work includes Wiebe and Rapaport ( 1988 ) , Hearst ( 1992 ) , Sack ( 1994 ) , and Wiebe ( 1994 ) ; see Esuli ( 2006 ) for an active bibliography ) .

First , it has been noted that in many natural language applications it is sufficient to use shallow parsing information ; information such as noun phrases ( NPs ) and other syntactic sequences have been found useful in many large-scale language processing applications including information extraction and text summarization ( Grishman , 1995 ; Appelt et al. , 1993 ) .

In addition , a fully flexible access system allows the retrieval of dictionary entries on the basis of constraints specifying any combination of phonetic , lexical , syntactic , and semantic information ( Boguraev et al. , 1987 ) .

Clarkson and Robinson ( 1997 ) developed a way of incorporating standard n-grams into the cache model , using mixtures of language models and also exponentially decaying the weight for the cache prediction depending on the recency of the word 's last

Riehemann 1993 ; Oliva 1994 ; Frank 1994 ; Opalka 1995 ; Sanfilippo 1995 ) .

Machine learning methods should be interchangeable : Transformation-based learning ( TBL ) ( Brill , 1993 ) and Memory-based learning ( MBL ) ( Daelemans et al. , 2002 ) have been applied to many different problems , so a single interchangeable component should be used to represent each method .

Recently , several alternative , often quite sophisticated approaches to collective classification have been proposed ( Neville and Jensen , 2000 ; Lafferty et al. , 2001 ; Getoor et al. , 2002 ; Taskar et al. , 2002 ; Taskar et al. , 2003 ; Taskar et al. , 2004 ; McCallum and Wellner , 2004 ) .

We tested the classification of verbs into semantic types using a verb list of 139 pre-classified items drawn from the lists published in Rosenbaum ( 1967 ) and Stockwell et al. ( 1973 ) .

The combination of likelihood and prior modeling , HMMs , and Viterbi decoding is fundamentally the same as the standard probabilistic approaches to speech recognition ( Bahl , Jelinek , and Mercer 1983 ) and tagging ( Church 1988 ) .

Cross-lingual Textual Entailment ( CLTE ) has been proposed by ( Mehdad et al. , 2010 ) as an extension of Textual Entailment ( Dagan and Glickman , 2004 ) that consists in deciding , given two texts T and H in different languages , if the meaning of H can be inferred from the meaning of T .

The EDR has close ties to the named entity recognition ( NER ) and coreference resolution tasks , which have been the focus of several recent investigations ( Bikel et al. , 1997 ; Miller et al. , 1998 ; Borthwick , 1999 ; Mikheev et al. , 1999 ; Soon et al. , 2001 ; Ng and Cardie , 2002 ; Florian et al. , 2004 ) , and have been at the center of evaluations such as : MUC-6 , MUC-7 , and the CoNLL '02 and CoNLL '03 shared tasks .

Other solutions such as complete caching of the corpora are not typically adopted due to legal concerns over copyright and redistribution of web data , issues considered at length by Fletcher ( 2004a ) .

This includes work on generalized expectation ( Mann and McCallum , 2010 ) , posterior regularization ( Ganchev et al. , 2010 ) and constraint driven learning ( Chang et al. , 2007 ; Chang et al. , 2010 ) .

A very similar formulation , for another grammar transformation , is given in Nederhof ( 1998 ) .

We run GIZA + + ( Och and Ney , 2000 ) on the training corpus in both directions ( Koehn et al. , 2003 ) to obtain the word alignment for each sentence pair .

Other attempts to address efficiency include the fast Transformation Based Learning ( TBL ) Toolkit ( Ngai and Florian , 2001 ) which dramatically speeds up training TBL systems , and the translation of TBL rules into finite state machines for very fast tagging ( Roche and Schabes , 1997 ) .

Consequently , fusion has been applied to a wide variety of pattern recognition and decision theoretic problems -- using a plethora of theories , techniques , and tools -- including some applications in computational linguistics ( e.g. , Brill and Wu 1998 ; van Halteren , Zavrel , and Daelemans 1998 ) and speech technology ( e.g. , Bowles and Damper 1989 ; Romary and Pierre11989 ) .

This observation has led some researchers , e.g. , Cooper and Paccia-Cooper ( 1980 ) , to claim a direct mapping between the syntactic phrase and the prosodic phrase .

These features are carefully designed to reduce the data sparseness problem and some of them are inspired by previous work ( He et al. , 2008 ; Gimpel and Smith , 2008 ; Marton and Resnik , 2008 ; Chiang et al. , 2009 ; Setiawan et al. , 2009 ; Shen et al. , 2009 ; Xiong et al. , 2009 ) : 1 .

In our previous work ( Zhang and Chai , 2009 ) , conversation entailment is formulated as the following : given a conversation segment D which is represented by a set of clauses D = d1 ∧ ... ∧ dm , and a hypothesis H represented by another set of clauses H = h1 ∧ ... ∧ hn , the prediction on whether D entails H is determined by the product of probabilities that each hypothesis clause hj is entailed from all the conversation segment clauses d1 ... dm as follows .

Notable early papers on graph-based semisupervised learning include Blum and Chawla ( 2001 ) , Bansal et al. ( 2002 ) , Kondor and Lafferty ( 2002 ) , and Joachims ( 2003 ) .

In addition , we consider several types of lexical features ( LexF ) inspired by previous work on agreement and disagreement ( Galley et al. , 2004 ; Misra and Walker , 2013 ) .

Berger et al. ( 2000 ) compared two retrieval approaches ( TF.IDF and query expansion ) and two predictive approaches ( statistical translation and latent variable models ) .

Others provide automatic mappings of natural language instructions to executable actions , such as interpreting navigation directions ( Chen and Mooney , 2011 ) or robot commands ( Tellex et al. , 2011 ; Matuszek et al. , 2012 ) .

The dialogue state is represented by a cumulative answer analysis which tracks , over multiple turns , the correct , incorrect , and not-yet-mentioned parts 1Other factors such as student confidence could be considered as well ( Callaway et al. , 2007 ) .

As noted above , it is well documented ( Roland and Jurafsky 1998 ) that subcategorization frames ( and their frequencies ) vary across domains .

Chen and Vijay-Shanker ( 2000 ) explore a number of related approaches to the extraction of a lexicalized TAG from the Penn-II Treebank with the aim of constructing a statistical model for parsing .

At the same time , we believe our method has advantages over the approach developed initially at IBM ( Brown et al. 1990 ; Brown et al. 1993 ) for training translation systems automatically .

Building on the work of Ruch et al. ( 2003 ) in the same domain , we present a generative approach that attempts to directly model the discourse structure of MEDLINE abstracts using Hidden Markov Models ( HMMs ) ; cfXXX ( Barzilay and Lee , 2004 ) .

The basic Python reflection has already been implemented and used for large scale experiments with POS tagging , using pyMPI ( a message passing interface library for Python ) to coordinate experiments across a cluster of over 100 machines ( Curran and Clark , 2003 ; Clark et al. , 2003 ) .

As an alternative , we rely on PubMed to retrieve an initial set of hits that we then postprocess in greater detail -- this is the standard pipeline architecture commonly employed in other question-answering systems ( Voorhees and Tice 1999 ; Hirschman and Gaizauskas 2001 ) .

The language grounding problem has received significant attention in recent years , owed in part to the wide availability of data sets ( e.g. Flickr , Von Ahn ( 2006 ) ) , computing power , improved computer vision models ( Oliva and Torralba , 2001 ; Lowe , 2004 ; Farhadi et al. , 2009 ; Parikh and Grauman , 2011 ) and neurological evidence of ties between the language , perceptual and motor systems in the brain ( Pulverm ¨ uller et al. , 2005 ; Tettamanti et al. , 2005 ; Aziz-Zadeh et al. , 2006 ) .

Although in this paper we take modus ponens as the main rule of inference , in general one can consider deductive closures with respect to weaker , nonstandard logics , ( cfXXX Levesque 1984 ; Frisch 1987 ; Patel-Schneider 1985 ) .

As Alshawi ( 1987 ) points out , given that no situations were envisaged where the information from the tape would be altered once installed in secondary storage , this simple and convenComputational Linguistics , Volume 13 , Numbers 3-4 , July-December 1987 205 Bran Boguraev and Ted Briscoe Large Lexicons for Natural Language Processing tional access strategy is perfectly adequate .

Other factors , such as the role of focus ( Grosz 1977 , 1978 ; Sidner 1983 ) or quantifier scoping ( Webber 1983 ) must play a role , too .

This paper describes an approach for sharing resources in various grammar formalisms such as Feature-Based Lexicalized Tree Adjoining Grammar ( FB-LTAG1 ) ( Vijay-Shanker , 1987 ; Vijay-Shanker and Joshi , 1988 ) and Head-Driven Phrase Structure Grammar ( HPSG ) ( Pollard and Sag , 1994 ) by a method of grammar conversion .