input
stringlengths
37
565
output
dict
schema
listlengths
0
0
Consequently , the proposed approach can be applied to different domains easily and it is especially useful for resource-limited domains .
{ "relations": { "used for": [ { "head": { "text": "it", "start": 84, "end": 86 }, "tail": { "text": "resource-limited domains", "start": 112, "end": 136 } } ] } }
[]
Evaluations conducted on two different domains for Chinese term extraction show significant improvements over existing techniques which verifies its efficiency and domain independent nature .
{ "relations": { "evaluate for": [ { "head": { "text": "Evaluations", "start": 0, "end": 11 }, "tail": { "text": "Chinese term extraction", "start": 51, "end": 74 } } ] } }
[]
Experiments on new term extraction indicate that the proposed approach can also serve as an effective tool for domain lexicon expansion .
{ "relations": { "used for": [ { "head": { "text": "approach", "start": 62, "end": 70 }, "tail": { "text": "domain lexicon expansion", "start": 111, "end": 135 } } ] } }
[]
We describe a method for identifying systematic patterns in translation data using part-of-speech tag sequences .
{ "relations": { "used for": [ { "head": { "text": "method", "start": 14, "end": 20 }, "tail": { "text": "systematic patterns in translation data", "start": 37, "end": 76 } }, { "head": { ...
[]
We incorporate this analysis into a diagnostic tool intended for developers of machine translation systems , and demonstrate how our application can be used by developers to explore patterns in machine translation output .
{ "relations": { "part of": [ { "head": { "text": "analysis", "start": 20, "end": 28 }, "tail": { "text": "diagnostic tool", "start": 36, "end": 51 } } ], "used for": [ { "head": { ...
[]
We study the number of hidden layers required by a multilayer neu-ral network with threshold units to compute a function f from n d to -LCB- O , I -RCB- .
{ "relations": { "used for": [ { "head": { "text": "number of hidden layers", "start": 13, "end": 36 }, "tail": { "text": "multilayer neu-ral network", "start": 51, "end": 77 } }, { "head": { ...
[]
We show that adding these conditions to Gib-son 's assumptions is not sufficient to ensure global computability with one hidden layer , by exhibiting a new non-local configuration , the `` critical cycle '' , which implies that f is not computable with one hidden layer .
{ "relations": { "hyponym of": [ { "head": { "text": "`` critical cycle ''", "start": 186, "end": 206 }, "tail": { "text": "non-local configuration", "start": 156, "end": 179 } } ] } }
[]
This paper presents an approach to estimate the intrinsic texture properties -LRB- albedo , shading , normal -RRB- of scenes from multiple view acquisition under unknown illumination conditions .
{ "relations": { "used for": [ { "head": { "text": "approach", "start": 23, "end": 31 }, "tail": { "text": "intrinsic texture properties -LRB- albedo , shading , normal -RRB- of scenes", "start": 48, "end": 124 } ...
[]
Unlike previous video relighting methods , the approach does not assume regions of uniform albedo , which makes it applicable to richly textured scenes .
{ "relations": { "compare": [ { "head": { "text": "approach", "start": 47, "end": 55 }, "tail": { "text": "video relighting methods", "start": 16, "end": 40 } } ], "used for": [ { "head": ...
[]
We show that intrinsic image methods can be used to refine an initial , low-frequency shading estimate based on a global lighting reconstruction from an original texture and coarse scene geometry in order to resolve the inherent global ambiguity in shading .
{ "relations": { "used for": [ { "head": { "text": "intrinsic image methods", "start": 13, "end": 36 }, "tail": { "text": "initial , low-frequency shading estimate", "start": 62, "end": 102 } }, { ...
[]
The method is applied to relight-ing of free-viewpoint rendering from multiple view video capture .
{ "relations": { "used for": [ { "head": { "text": "method", "start": 4, "end": 10 }, "tail": { "text": "relight-ing of free-viewpoint rendering", "start": 25, "end": 64 } }, { "head": { ...
[]
This demonstrates relighting with reproduction of fine surface detail .
{ "relations": { "feature of": [ { "head": { "text": "reproduction of fine surface detail", "start": 34, "end": 69 }, "tail": { "text": "relighting", "start": 18, "end": 28 } } ] } }
[]
Following recent developments in the automatic evaluation of machine translation and document summarization , we present a similar approach , implemented in a measure called POURPRE , for automatically evaluating answers to definition questions .
{ "relations": { "evaluate for": [ { "head": { "text": "automatic evaluation", "start": 37, "end": 57 }, "tail": { "text": "machine translation", "start": 61, "end": 80 } }, { "head": { ...
[]
Experiments with the TREC 2003 and TREC 2004 QA tracks indicate that rankings produced by our metric correlate highly with official rankings , and that POURPRE outperforms direct application of existing metrics .
{ "relations": { "evaluate for": [ { "head": { "text": "TREC 2003 and TREC 2004 QA tracks", "start": 21, "end": 54 }, "tail": { "text": "metric", "start": 94, "end": 100 } }, { "head": { ...
[]
Recent advances in Automatic Speech Recognition technology have put the goal of naturally sounding dialog systems within reach .
{ "relations": { "used for": [ { "head": { "text": "Automatic Speech Recognition technology", "start": 19, "end": 58 }, "tail": { "text": "dialog systems", "start": 99, "end": 113 } } ] } }
[]
The issue of system response to users has been extensively studied by the natural language generation community , though rarely in the context of dialog systems .
{ "relations": { "part of": [ { "head": { "text": "system response", "start": 13, "end": 28 }, "tail": { "text": "natural language generation community", "start": 74, "end": 111 } } ], "compare": [ ...
[]
We show how research in generation can be adapted to dialog systems , and how the high cost of hand-crafting knowledge-based generation systems can be overcome by employing machine learning techniques .
{ "relations": { "used for": [ { "head": { "text": "generation", "start": 24, "end": 34 }, "tail": { "text": "dialog systems", "start": 53, "end": 67 } }, { "head": { "text": "machine le...
[]
We present a tool , called ILIMP , which takes as input a raw text in French and produces as output the same text in which every occurrence of the pronoun il is tagged either with tag -LSB- ANA -RSB- for anaphoric or -LSB- IMP -RSB- for impersonal or expletive .
{ "relations": { "used for": [ { "head": { "text": "raw text in French", "start": 58, "end": 76 }, "tail": { "text": "tool", "start": 13, "end": 17 } } ] } }
[]
This tool is therefore designed to distinguish between the anaphoric occurrences of il , for which an anaphora resolution system has to look for an antecedent , and the expletive occurrences of this pronoun , for which it does not make sense to look for an antecedent .
{ "relations": { "used for": [ { "head": { "text": "tool", "start": 5, "end": 9 }, "tail": { "text": "anaphoric occurrences of il", "start": 59, "end": 86 } }, { "head": { "text": "anaph...
[]
The precision rate for ILIMP is 97,5 % .
{ "relations": { "evaluate for": [ { "head": { "text": "precision rate", "start": 4, "end": 18 }, "tail": { "text": "ILIMP", "start": 23, "end": 28 } } ] } }
[]
Other tasks using the method developed for ILIMP are described briefly , as well as the use of ILIMP in a modular syntactic analysis system .
{ "relations": { "used for": [ { "head": { "text": "method", "start": 22, "end": 28 }, "tail": { "text": "tasks", "start": 6, "end": 11 } }, { "head": { "text": "method", "star...
[]
Little is thus known about the robustness of speech cues in the wild .
{ "relations": { "feature of": [ { "head": { "text": "speech cues", "start": 45, "end": 56 }, "tail": { "text": "robustness", "start": 31, "end": 41 } } ] } }
[]
This study compares the effect of noise and reverberation on depression prediction using 1 -RRB- standard mel-frequency cepstral coefficients -LRB- MFCCs -RRB- , and 2 -RRB- features designed for noise robustness , damped oscillator cepstral coefficients -LRB- DOCCs -RRB- .
{ "relations": { "conjunction": [ { "head": { "text": "noise", "start": 34, "end": 39 }, "tail": { "text": "reverberation", "start": 44, "end": 57 } }, { "head": { "text": "mel-frequency...
[]
Results using additive noise and reverberation reveal a consistent pattern of findings for multiple evaluation metrics under both matched and mismatched conditions .
{ "relations": { "conjunction": [ { "head": { "text": "additive noise", "start": 14, "end": 28 }, "tail": { "text": "reverberation", "start": 33, "end": 46 } } ] } }
[]
First and most notably : standard MFCC features suffer dramatically under test/train mismatch for both noise and reverberation ; DOCC features are far more robust .
{ "relations": { "conjunction": [ { "head": { "text": "noise", "start": 103, "end": 108 }, "tail": { "text": "reverberation", "start": 113, "end": 126 } } ], "compare": [ { "head": { ...
[]
Third , artificial neural networks tend to outperform support vector regression .
{ "relations": { "compare": [ { "head": { "text": "artificial neural networks", "start": 8, "end": 34 }, "tail": { "text": "support vector regression", "start": 54, "end": 79 } } ] } }
[]
Fourth , spontaneous speech appears to offer better robustness than read speech .
{ "relations": { "compare": [ { "head": { "text": "spontaneous speech", "start": 9, "end": 27 }, "tail": { "text": "read speech", "start": 68, "end": 79 } } ] } }
[]
Finally , a cross-corpus -LRB- and cross-language -RRB- experiment reveals better noise and reverberation robustness for DOCCs than for MFCCs .
{ "relations": { "evaluate for": [ { "head": { "text": "cross-corpus -LRB- and cross-language -RRB- experiment", "start": 12, "end": 66 }, "tail": { "text": "DOCCs", "start": 121, "end": 126 } }, { ...
[]
This paper proposes document oriented preference sets -LRB- DoPS -RRB- for the disambiguation of the dependency structure of sentences .
{ "relations": { "used for": [ { "head": { "text": "document oriented preference sets -LRB- DoPS -RRB-", "start": 20, "end": 70 }, "tail": { "text": "disambiguation of the dependency structure", "start": 79, "end": 121 ...
[]
Sentence ambiguities can be resolved by using domain targeted preference knowledge without using complicated large knowledgebases .
{ "relations": { "used for": [ { "head": { "text": "domain targeted preference knowledge", "start": 46, "end": 82 }, "tail": { "text": "Sentence ambiguities", "start": 0, "end": 20 } } ], "compare": [ ...
[]
Implementation and empirical results are described for the the analysis of dependency structures of Japanese patent claim sentences .
{ "relations": { "feature of": [ { "head": { "text": "dependency structures", "start": 75, "end": 96 }, "tail": { "text": "Japanese patent claim sentences", "start": 100, "end": 131 } } ] } }
[]
Multimodal interfaces require effective parsing and understanding of utterances whose content is distributed across multiple input modes .
{ "relations": { "used for": [ { "head": { "text": "parsing", "start": 40, "end": 47 }, "tail": { "text": "Multimodal interfaces", "start": 0, "end": 21 } } ] } }
[]
Johnston 1998 presents an approach in which strategies for multimodal integration are stated declaratively using a unification-based grammar that is used by a multidimensional chart parser to compose inputs .
{ "relations": { "used for": [ { "head": { "text": "approach", "start": 26, "end": 34 }, "tail": { "text": "multimodal integration", "start": 59, "end": 81 } }, { "head": { "text": "unif...
[]
In this paper , we present an alternative approach in which multimodal parsing and understanding are achieved using a weighted finite-state device which takes speech and gesture streams as inputs and outputs their joint interpretation .
{ "relations": { "used for": [ { "head": { "text": "approach", "start": 42, "end": 50 }, "tail": { "text": "multimodal parsing and understanding", "start": 60, "end": 96 } }, { "head": { ...
[]
This approach is significantly more efficient , enables tight-coupling of multimodal understanding with speech recognition , and provides a general probabilistic framework for multimodal ambiguity resolution .
{ "relations": { "used for": [ { "head": { "text": "approach", "start": 5, "end": 13 }, "tail": { "text": "multimodal ambiguity resolution", "start": 176, "end": 207 } } ], "conjunction": [ { ...
[]
Recently , we initiated a project to develop a phonetically-based spoken language understanding system called SUMMIT .
{ "relations": { "hyponym of": [ { "head": { "text": "SUMMIT", "start": 110, "end": 116 }, "tail": { "text": "phonetically-based spoken language understanding system", "start": 47, "end": 102 } } ] } }
[]
In contrast to many of the past efforts that make use of heuristic rules whose development requires intense knowledge engineering , our approach attempts to express the speech knowledge within a formal framework using well-defined mathematical tools .
{ "relations": { "used for": [ { "head": { "text": "knowledge engineering", "start": 108, "end": 129 }, "tail": { "text": "heuristic rules", "start": 57, "end": 72 } }, { "head": { "text...
[]
In our system , features and decision strategies are discovered and trained automatically , using a large body of speech data .
{ "relations": { "conjunction": [ { "head": { "text": "features", "start": 16, "end": 24 }, "tail": { "text": "decision strategies", "start": 29, "end": 48 } } ], "used for": [ { "head": {...
[]
This paper describes an implemented program that takes a tagged text corpus and generates a partial list of the subcategorization frames in which each verb occurs .
{ "relations": { "evaluate for": [ { "head": { "text": "tagged text corpus", "start": 57, "end": 75 }, "tail": { "text": "program", "start": 36, "end": 43 } } ] } }
[]
We present a method for estimating the relative pose of two calibrated or uncalibrated non-overlapping surveillance cameras from observing a moving object .
{ "relations": { "used for": [ { "head": { "text": "method", "start": 13, "end": 19 }, "tail": { "text": "relative pose of two calibrated or uncalibrated non-overlapping surveillance cameras", "start": 39, "end": 123 ...
[]
We show how to tackle the problem of missing point correspondences heavily required by SfM pipelines and how to go beyond this basic paradigm .
{ "relations": { "used for": [ { "head": { "text": "SfM pipelines", "start": 87, "end": 100 }, "tail": { "text": "missing point correspondences", "start": 37, "end": 66 } } ] } }
[]
We relax the non-linear nature of the problem by accepting two assumptions which surveillance scenarios offer , ie .
{ "relations": { "feature of": [ { "head": { "text": "non-linear nature", "start": 13, "end": 30 }, "tail": { "text": "problem", "start": 38, "end": 45 } } ] } }
[]
By those assumptions we cast the problem as a Quadratic Eigenvalue Problem offering an elegant way of treating nonlinear monomials and delivering a quasi closed-form solution as a reliable starting point for a further bundle adjustment .
{ "relations": { "used for": [ { "head": { "text": "Quadratic Eigenvalue Problem", "start": 46, "end": 74 }, "tail": { "text": "problem", "start": 33, "end": 40 } }, { "head": { "text": ...
[]
We are the first to bring the closed form solution to such a very practical problem arising in video surveillance .
{ "relations": { "used for": [ { "head": { "text": "closed form solution", "start": 30, "end": 50 }, "tail": { "text": "problem", "start": 76, "end": 83 } } ], "feature of": [ { "head": { ...
[]
In this paper , we propose a human action recognition system suitable for embedded computer vision applications in security systems , human-computer interaction and intelligent environments .
{ "relations": { "used for": [ { "head": { "text": "human action recognition system", "start": 29, "end": 60 }, "tail": { "text": "embedded computer vision applications", "start": 74, "end": 111 } }, { ...
[]
Our system is suitable for embedded computer vision application based on three reasons .
{ "relations": { "used for": [ { "head": { "text": "system", "start": 4, "end": 10 }, "tail": { "text": "embedded computer vision application", "start": 27, "end": 63 } } ] } }
[]
Firstly , the system was based on a linear Support Vector Machine -LRB- SVM -RRB- classifier where classification progress can be implemented easily and quickly in embedded hardware .
{ "relations": { "used for": [ { "head": { "text": "linear Support Vector Machine -LRB- SVM -RRB- classifier", "start": 36, "end": 92 }, "tail": { "text": "system", "start": 14, "end": 20 } }, { "...
[]
Secondly , we use compacted motion features easily obtained from videos .
{ "relations": { "used for": [ { "head": { "text": "videos", "start": 65, "end": 71 }, "tail": { "text": "compacted motion features", "start": 18, "end": 43 } } ] } }
[]
We address the limitations of the well known Motion History Image -LRB- MHI -RRB- and propose a new Hierarchical Motion History Histogram -LRB- HMHH -RRB- feature to represent the motion information .
{ "relations": { "used for": [ { "head": { "text": "Hierarchical Motion History Histogram -LRB- HMHH -RRB- feature", "start": 100, "end": 162 }, "tail": { "text": "motion information", "start": 180, "end": 198 } ...
[]
HMHH not only provides rich motion information , but also remains computationally inexpensive .
{ "relations": { "used for": [ { "head": { "text": "HMHH", "start": 0, "end": 4 }, "tail": { "text": "rich motion information", "start": 23, "end": 46 } } ] } }
[]
Finally , we combine MHI and HMHH together and extract a low dimension feature vector to be used in the SVM classifiers .
{ "relations": { "conjunction": [ { "head": { "text": "MHI", "start": 21, "end": 24 }, "tail": { "text": "HMHH", "start": 29, "end": 33 } } ], "used for": [ { "head": { "text": "...
[]
Experimental results show that our system achieves significant improvement on the recognition performance .
{ "relations": { "evaluate for": [ { "head": { "text": "recognition", "start": 82, "end": 93 }, "tail": { "text": "system", "start": 35, "end": 41 } } ] } }
[]
In this paper I will argue for a model of grammatical processing that is based on uniform processing and knowledge sources .
{ "relations": { "used for": [ { "head": { "text": "uniform processing", "start": 82, "end": 100 }, "tail": { "text": "model of grammatical processing", "start": 33, "end": 64 } }, { "head": { ...
[]
The main feature of this model is to view parsing and generation as two strongly interleaved tasks performed by a single parametrized deduction process .
{ "relations": { "conjunction": [ { "head": { "text": "parsing", "start": 42, "end": 49 }, "tail": { "text": "generation", "start": 54, "end": 64 } } ], "hyponym of": [ { "head": { ...
[]
Link detection has been regarded as a core technology for the Topic Detection and Tracking tasks of new event detection .
{ "relations": { "used for": [ { "head": { "text": "Link detection", "start": 0, "end": 14 }, "tail": { "text": "Topic Detection and Tracking tasks of new event detection", "start": 62, "end": 119 } } ] }...
[]
In this paper we formulate story link detection and new event detection as information retrieval task and hypothesize on the impact of precision and recall on both systems .
{ "relations": { "conjunction": [ { "head": { "text": "story link detection", "start": 27, "end": 47 }, "tail": { "text": "new event detection", "start": 52, "end": 71 } }, { "head": { "...
[]
Motivated by these arguments , we introduce a number of new performance enhancing techniques including part of speech tagging , new similarity measures and expanded stop lists .
{ "relations": { "part of": [ { "head": { "text": "part of speech tagging", "start": 103, "end": 125 }, "tail": { "text": "performance enhancing techniques", "start": 60, "end": 92 } }, { "head": ...
[]
We attempt to understand visual classification in humans using both psy-chophysical and machine learning techniques .
{ "relations": { "used for": [ { "head": { "text": "psy-chophysical and machine learning techniques", "start": 68, "end": 115 }, "tail": { "text": "visual classification", "start": 25, "end": 46 } } ] } }
[]
Frontal views of human faces were used for a gender classification task .
{ "relations": { "used for": [ { "head": { "text": "Frontal views of human faces", "start": 0, "end": 28 }, "tail": { "text": "gender classification task", "start": 45, "end": 71 } } ] } }
[]
Several hyperplane learning algorithms were used on the same classification task using the Principal Components of the texture and flowfield representation of the faces .
{ "relations": { "used for": [ { "head": { "text": "hyperplane learning algorithms", "start": 8, "end": 38 }, "tail": { "text": "classification task", "start": 61, "end": 80 } }, { "head": { ...
[]
The classification performance of the learning algorithms was estimated using the face database with the true gender of the faces as labels , and also with the gender estimated by the subjects .
{ "relations": { "used for": [ { "head": { "text": "learning algorithms", "start": 38, "end": 57 }, "tail": { "text": "classification", "start": 4, "end": 18 } } ], "evaluate for": [ { "he...
[]
Our results suggest that human classification can be modeled by some hyperplane algorithms in the feature space we used .
{ "relations": { "used for": [ { "head": { "text": "hyperplane algorithms", "start": 69, "end": 90 }, "tail": { "text": "human classification", "start": 25, "end": 45 } } ], "feature of": [ { ...
[]
For classification , the brain needs more processing for stimuli close to that hyperplane than for those further away .
{ "relations": { "compare": [ { "head": { "text": "hyperplane", "start": 79, "end": 89 }, "tail": { "text": "those", "start": 99, "end": 104 } } ] } }
[]
In this paper , we present a corpus-based supervised word sense disambiguation -LRB- WSD -RRB- system for Dutch which combines statistical classification -LRB- maximum entropy -RRB- with linguistic information .
{ "relations": { "used for": [ { "head": { "text": "corpus-based supervised word sense disambiguation -LRB- WSD -RRB- system", "start": 29, "end": 101 }, "tail": { "text": "Dutch", "start": 106, "end": 111 } } ...
[]
Instead of building individual classifiers per ambiguous wordform , we introduce a lemma-based approach .
{ "relations": { "compare": [ { "head": { "text": "classifiers", "start": 31, "end": 42 }, "tail": { "text": "lemma-based approach", "start": 83, "end": 103 } } ], "used for": [ { "head": ...
[]
The advantage of this novel method is that it clusters all inflected forms of an ambiguous word in one classifier , therefore augmenting the training material available to the algorithm .
{ "relations": { "feature of": [ { "head": { "text": "inflected forms", "start": 59, "end": 74 }, "tail": { "text": "ambiguous word", "start": 81, "end": 95 } } ] } }
[]
Testing the lemma-based model on the Dutch Senseval-2 test data , we achieve a significant increase in accuracy over the wordform model .
{ "relations": { "compare": [ { "head": { "text": "lemma-based model", "start": 12, "end": 29 }, "tail": { "text": "wordform model", "start": 121, "end": 135 } } ], "evaluate for": [ { "he...
[]
We propose an exact , general and efficient coarse-to-fine energy minimization strategy for semantic video segmenta-tion .
{ "relations": { "used for": [ { "head": { "text": "coarse-to-fine energy minimization strategy", "start": 44, "end": 87 }, "tail": { "text": "semantic video segmenta-tion", "start": 92, "end": 120 } } ] ...
[]
Our strategy is based on a hierarchical abstraction of the supervoxel graph that allows us to minimize an energy defined at the finest level of the hierarchy by minimizing a series of simpler energies defined over coarser graphs .
{ "relations": { "used for": [ { "head": { "text": "hierarchical abstraction of the supervoxel graph", "start": 27, "end": 75 }, "tail": { "text": "strategy", "start": 4, "end": 12 } } ] } }
[]
It is general , i.e. , it can be used to minimize any energy function -LRB- e.g. , unary , pairwise , and higher-order terms -RRB- with any existing energy minimization algorithm -LRB- e.g. , graph cuts and belief propagation -RRB- .
{ "relations": { "used for": [ { "head": { "text": "it", "start": 23, "end": 25 }, "tail": { "text": "energy function", "start": 54, "end": 69 } }, { "head": { "text": "energy minimizati...
[]
It also gives significant speedups in inference for several datasets with varying degrees of spatio-temporal continuity .
{ "relations": { "used for": [ { "head": { "text": "It", "start": 0, "end": 2 }, "tail": { "text": "inference", "start": 38, "end": 47 } } ], "evaluate for": [ { "head": { "text"...
[]
We also discuss the strengths and weaknesses of our strategy relative to existing hierarchical approaches , and the kinds of image and video data that provide the best speedups .
{ "relations": { "compare": [ { "head": { "text": "strategy", "start": 52, "end": 60 }, "tail": { "text": "hierarchical approaches", "start": 82, "end": 105 } } ] } }
[]
Motivated by the success of ensemble methods in machine learning and other areas of natural language processing , we developed a multi-strategy and multi-source approach to question answering which is based on combining the results from different answering agents searching for answers in multiple corpora .
{ "relations": { "used for": [ { "head": { "text": "ensemble methods", "start": 28, "end": 44 }, "tail": { "text": "machine learning", "start": 48, "end": 64 } }, { "head": { "text": "en...
[]
The answering agents adopt fundamentally different strategies , one utilizing primarily knowledge-based mechanisms and the other adopting statistical techniques .
{ "relations": { "used for": [ { "head": { "text": "strategies", "start": 51, "end": 61 }, "tail": { "text": "answering agents", "start": 4, "end": 20 } }, { "head": { "text": "knowledge...
[]
We present our multi-level answer resolution algorithm that combines results from the answering agents at the question , passage , and/or answer levels .
{ "relations": { "used for": [ { "head": { "text": "answering agents", "start": 86, "end": 102 }, "tail": { "text": "multi-level answer resolution algorithm", "start": 15, "end": 54 } } ] } }
[]
Experiments evaluating the effectiveness of our answer resolution algorithm show a 35.0 % relative improvement over our baseline system in the number of questions correctly answered , and a 32.8 % improvement according to the average precision metric .
{ "relations": { "compare": [ { "head": { "text": "answer resolution algorithm", "start": 48, "end": 75 }, "tail": { "text": "baseline system", "start": 120, "end": 135 } } ], "evaluate for": [ { ...
[]
Word Identification has been an important and active issue in Chinese Natural Language Processing .
{ "relations": { "hyponym of": [ { "head": { "text": "Word Identification", "start": 0, "end": 19 }, "tail": { "text": "Chinese Natural Language Processing", "start": 62, "end": 97 } } ] } }
[]
In this paper , a new mechanism , based on the concept of sublanguage , is proposed for identifying unknown words , especially personal names , in Chinese newspapers .
{ "relations": { "used for": [ { "head": { "text": "mechanism", "start": 22, "end": 31 }, "tail": { "text": "unknown words", "start": 100, "end": 113 } }, { "head": { "text": "sublanguag...
[]
The proposed mechanism includes title-driven name recognition , adaptive dynamic word formation , identification of 2-character and 3-character Chinese names without title .
{ "relations": { "part of": [ { "head": { "text": "title-driven name recognition", "start": 32, "end": 61 }, "tail": { "text": "mechanism", "start": 13, "end": 22 } }, { "head": { "text"...
[]
This report describes Paul , a computer text generation system designed to create cohesive text through the use of lexical substitutions .
{ "relations": { "hyponym of": [ { "head": { "text": "Paul", "start": 22, "end": 26 }, "tail": { "text": "computer text generation system", "start": 31, "end": 62 } } ], "used for": [ { "h...
[]
Specifically , this system is designed to deterministically choose between pronominalization , superordinate substitution , and definite noun phrase reiteration .
{ "relations": { "compare": [ { "head": { "text": "pronominalization", "start": 75, "end": 92 }, "tail": { "text": "superordinate substitution", "start": 95, "end": 121 } }, { "head": { ...
[]
The system identifies a strength of antecedence recovery for each of the lexical substitutions .
{ "relations": { "used for": [ { "head": { "text": "system", "start": 4, "end": 10 }, "tail": { "text": "antecedence recovery", "start": 36, "end": 56 } }, { "head": { "text": "anteceden...
[]
It describes the automated training and evaluation of an Optimal Position Policy , a method of locating the likely positions of topic-bearing sentences based on genre-specific regularities of discourse structure .
{ "relations": { "used for": [ { "head": { "text": "method", "start": 85, "end": 91 }, "tail": { "text": "positions of topic-bearing sentences", "start": 115, "end": 151 } }, { "head": { ...
[]
This method can be used in applications such as information retrieval , routing , and text summarization .
{ "relations": { "used for": [ { "head": { "text": "method", "start": 5, "end": 11 }, "tail": { "text": "applications", "start": 27, "end": 39 } } ], "hyponym of": [ { "head": { ...
[]
We describe a general framework for online multiclass learning based on the notion of hypothesis sharing .
{ "relations": { "used for": [ { "head": { "text": "framework", "start": 22, "end": 31 }, "tail": { "text": "online multiclass learning", "start": 36, "end": 62 } }, { "head": { "text": ...
[]
We generalize the multiclass Perceptron to our framework and derive a unifying mistake bound analysis .
{ "relations": { "used for": [ { "head": { "text": "multiclass Perceptron", "start": 18, "end": 39 }, "tail": { "text": "framework", "start": 47, "end": 56 } } ] } }
[]
We demonstrate the merits of our approach by comparing it to previous methods on both synthetic and natural datasets .
{ "relations": { "compare": [ { "head": { "text": "it", "start": 22, "end": 24 }, "tail": { "text": "methods", "start": 70, "end": 77 } } ], "evaluate for": [ { "head": { "text":...
[]
We describe a set of supervised machine learning experiments centering on the construction of statistical models of WH-questions .
{ "relations": { "used for": [ { "head": { "text": "supervised machine learning", "start": 21, "end": 48 }, "tail": { "text": "statistical models of WH-questions", "start": 94, "end": 128 } } ] } }
[]
These models , which are built from shallow linguistic features of questions , are employed to predict target variables which represent a user 's informational goals .
{ "relations": { "used for": [ { "head": { "text": "shallow linguistic features of questions", "start": 36, "end": 76 }, "tail": { "text": "models", "start": 6, "end": 12 } } ] } }
[]
We argue in favor of the the use of labeled directed graph to represent various types of linguistic structures , and illustrate how this allows one to view NLP tasks as graph transformations .
{ "relations": { "used for": [ { "head": { "text": "labeled directed graph", "start": 36, "end": 58 }, "tail": { "text": "linguistic structures", "start": 89, "end": 110 } }, { "head": { ...
[]
We present a general method for learning such transformations from an annotated corpus and describe experiments with two applications of the method : identification of non-local depenencies -LRB- using Penn Treebank data -RRB- and semantic role labeling -LRB- using Proposition Bank data -RRB- .
{ "relations": { "used for": [ { "head": { "text": "method", "start": 21, "end": 27 }, "tail": { "text": "transformations", "start": 46, "end": 61 } }, { "head": { "text": "annotated cor...
[]
We describe a generative probabilistic model of natural language , which we call HBG , that takes advantage of detailed linguistic information to resolve ambiguity .
{ "relations": { "used for": [ { "head": { "text": "linguistic information", "start": 120, "end": 142 }, "tail": { "text": "ambiguity", "start": 154, "end": 163 } } ] } }
[]
HBG incorporates lexical , syntactic , semantic , and structural information from the parse tree into the disambiguation process in a novel way .
{ "relations": { "used for": [ { "head": { "text": "HBG", "start": 0, "end": 3 }, "tail": { "text": "disambiguation process", "start": 106, "end": 128 } }, { "head": { "text": "lexical ,...
[]
We use a corpus of bracketed sentences , called a Treebank , in combination with decision tree building to tease out the relevant aspects of a parse tree that will determine the correct parse of a sentence .
{ "relations": { "conjunction": [ { "head": { "text": "corpus of bracketed sentences", "start": 9, "end": 38 }, "tail": { "text": "decision tree building", "start": 81, "end": 103 } } ], "used for": [ ...
[]
This stands in contrast to the usual approach of further grammar tailoring via the usual linguistic introspection in the hope of generating the correct parse .
{ "relations": { "used for": [ { "head": { "text": "grammar tailoring", "start": 57, "end": 74 }, "tail": { "text": "parse", "start": 152, "end": 157 } }, { "head": { "text": "linguistic...
[]
In head-to-head tests against one of the best existing robust probabilistic parsing models , which we call P-CFG , the HBG model significantly outperforms P-CFG , increasing the parsing accuracy rate from 60 % to 75 % , a 37 % reduction in error .
{ "relations": { "hyponym of": [ { "head": { "text": "P-CFG", "start": 107, "end": 112 }, "tail": { "text": "robust probabilistic parsing models", "start": 55, "end": 90 } } ], "compare": [ { ...
[]
The framework of the analysis is model-theoretic semantics .
{ "relations": { "used for": [ { "head": { "text": "model-theoretic semantics", "start": 33, "end": 58 }, "tail": { "text": "analysis", "start": 21, "end": 29 } } ] } }
[]
This paper addresses the issue of word-sense ambiguity in extraction from machine-readable resources for the construction of large-scale knowledge sources .
{ "relations": { "used for": [ { "head": { "text": "machine-readable resources", "start": 74, "end": 100 }, "tail": { "text": "word-sense ambiguity", "start": 34, "end": 54 } }, { "head": { ...
[]
We describe two experiments : one which ignored word-sense distinctions , resulting in 6.3 % accuracy for semantic classification of verbs based on -LRB- Levin , 1993 -RRB- ; and one which exploited word-sense distinctions , resulting in 97.9 % accuracy .
{ "relations": { "evaluate for": [ { "head": { "text": "accuracy", "start": 93, "end": 101 }, "tail": { "text": "semantic classification", "start": 106, "end": 129 } } ] } }
[]
These experiments were dual purpose : -LRB- 1 -RRB- to validate the central thesis of the work of -LRB- Levin , 1993 -RRB- , i.e. , that verb semantics and syntactic behavior are predictably related ; -LRB- 2 -RRB- to demonstrate that a 15-fold improvement can be achieved in deriving semantic information from syntactic...
{ "relations": { "conjunction": [ { "head": { "text": "verb semantics", "start": 137, "end": 151 }, "tail": { "text": "syntactic behavior", "start": 156, "end": 174 } } ], "used for": [ { ...
[]