input stringlengths 37 565 | output dict | schema listlengths 0 0 |
|---|---|---|
In this paper , we introduce a closed-form solution to systematically combine the limited training data with some generic qualitative knowledge for BN parameter learning . | {
"relations": {
"conjunction": [
{
"head": {
"text": "limited training data",
"start": 82,
"end": 103
},
"tail": {
"text": "qualitative knowledge",
"start": 122,
"end": 143
}
}
],
"used for": [
{... | [] |
To validate our method , we compare it with the Maximum Likelihood -LRB- ML -RRB- estimation method under sparse data and with the Expectation Maximization -LRB- EM -RRB- algorithm under incomplete data respectively . | {
"relations": {
"compare": [
{
"head": {
"text": "it",
"start": 40,
"end": 42
},
"tail": {
"text": "Maximum Likelihood -LRB- ML -RRB- estimation method",
"start": 48,
"end": 99
}
},
{
"head": {
... | [] |
To further demonstrate its applications for computer vision , we apply it to learn a BN model for facial Action Unit -LRB- AU -RRB- recognition from real image data . | {
"relations": {
"used for": [
{
"head": {
"text": "it",
"start": 23,
"end": 25
},
"tail": {
"text": "computer vision",
"start": 44,
"end": 59
}
},
{
"head": {
"text": "it",
"s... | [] |
The experimental results show that with simple and generic qualitative constraints and using only a small amount of training data , our method can robustly and accurately estimate the BN model parameters . | {
"relations": {
"conjunction": [
{
"head": {
"text": "generic qualitative constraints",
"start": 51,
"end": 82
},
"tail": {
"text": "training data",
"start": 116,
"end": 129
}
}
],
"used for": [
... | [] |
In this paper we introduce a modal language LT for imposing constraints on trees , and an extension LT -LRB- LF -RRB- for imposing constraints on trees decorated with feature structures . | {
"relations": {
"used for": [
{
"head": {
"text": "modal language LT",
"start": 29,
"end": 46
},
"tail": {
"text": "constraints on trees",
"start": 60,
"end": 80
}
},
{
"head": {
"text"... | [] |
The motivation for introducing these languages is to provide tools for formalising grammatical frameworks perspicuously , and the paper illustrates this by showing how the leading ideas of GPSG can be captured in LT -LRB- LF -RRB- . | {
"relations": {
"used for": [
{
"head": {
"text": "languages",
"start": 37,
"end": 46
},
"tail": {
"text": "grammatical frameworks",
"start": 83,
"end": 105
}
},
{
"head": {
"text": "GP... | [] |
Previous research has demonstrated the utility of clustering in inducing semantic verb classes from undisambiguated corpus data . | {
"relations": {
"used for": [
{
"head": {
"text": "clustering",
"start": 50,
"end": 60
},
"tail": {
"text": "inducing semantic verb classes",
"start": 64,
"end": 94
}
},
{
"head": {
"te... | [] |
We describe a new approach which involves clustering subcategorization frame -LRB- SCF -RRB- distributions using the Information Bottleneck and nearest neighbour methods . | {
"relations": {
"part of": [
{
"head": {
"text": "clustering subcategorization frame -LRB- SCF -RRB- distributions",
"start": 42,
"end": 106
},
"tail": {
"text": "approach",
"start": 18,
"end": 26
}
}
],
... | [] |
A novel evaluation scheme is proposed which accounts for the effect of polysemy on the clusters , offering us a good insight into the potential and limitations of semantically classifying undisambiguated SCF data . | {
"relations": {
"used for": [
{
"head": {
"text": "evaluation scheme",
"start": 8,
"end": 25
},
"tail": {
"text": "polysemy",
"start": 71,
"end": 79
}
}
],
"evaluate for": [
{
"head": {
... | [] |
Due to the capacity of pan-tilt-zoom -LRB- PTZ -RRB- cameras to simultaneously cover a panoramic area and maintain high resolution imagery , researches in automated surveillance systems with multiple PTZ cameras have become increasingly important . | {
"relations": {
"used for": [
{
"head": {
"text": "pan-tilt-zoom -LRB- PTZ -RRB- cameras",
"start": 23,
"end": 60
},
"tail": {
"text": "panoramic area",
"start": 87,
"end": 101
}
},
{
"head": {
... | [] |
Most existing algorithms require the prior knowledge of intrinsic parameters of the PTZ camera to infer the relative positioning and orientation among multiple PTZ cameras . | {
"relations": {
"used for": [
{
"head": {
"text": "algorithms",
"start": 14,
"end": 24
},
"tail": {
"text": "relative positioning",
"start": 108,
"end": 128
}
},
{
"head": {
"text": "al... | [] |
To overcome this limitation , we propose a novel mapping algorithm that derives the relative positioning and orientation between two PTZ cameras based on a unified polynomial model . | {
"relations": {
"used for": [
{
"head": {
"text": "mapping algorithm",
"start": 49,
"end": 66
},
"tail": {
"text": "relative positioning",
"start": 84,
"end": 104
}
},
{
"head": {
"text... | [] |
Experimental results demonstrate that our proposed algorithm presents substantially reduced computational complexity and improved flexibility at the cost of slightly decreased pixel accuracy , as compared with the work of Chen and Wang . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "computational complexity",
"start": 92,
"end": 116
},
"tail": {
"text": "algorithm",
"start": 51,
"end": 60
}
},
{
"head": {
"text... | [] |
This slightly decreased pixel accuracy can be compensated by consistent labeling approaches without added cost for the application of automated surveillance systems along with changing configurations and a larger number of PTZ cameras . | {
"relations": {
"used for": [
{
"head": {
"text": "consistent labeling approaches",
"start": 61,
"end": 91
},
"tail": {
"text": "pixel accuracy",
"start": 24,
"end": 38
}
}
]
}
} | [] |
This paper presents a new two-pass algorithm for Extra Large -LRB- more than 1M words -RRB- Vocabulary COntinuous Speech recognition based on the Information Retrieval -LRB- ELVIRCOS -RRB- . | {
"relations": {
"used for": [
{
"head": {
"text": "two-pass algorithm",
"start": 26,
"end": 44
},
"tail": {
"text": "Extra Large -LRB- more than 1M words -RRB- Vocabulary COntinuous Speech recognition",
"start": 49,
"end": ... | [] |
The principle of this approach is to decompose a recognition process into two passes where the first pass builds the words subset for the second pass recognition by using information retrieval procedure . | {
"relations": {
"hyponym of": [
{
"head": {
"text": "first pass",
"start": 95,
"end": 105
},
"tail": {
"text": "passes",
"start": 78,
"end": 84
}
},
{
"head": {
"text": "second pass rec... | [] |
Word graph composition for continuous speech is presented . | {
"relations": {
"used for": [
{
"head": {
"text": "Word graph composition",
"start": 0,
"end": 22
},
"tail": {
"text": "continuous speech",
"start": 27,
"end": 44
}
}
]
}
} | [] |
With this approach a high performances for large vocabulary speech recognition can be obtained . | {
"relations": {
"used for": [
{
"head": {
"text": "approach",
"start": 10,
"end": 18
},
"tail": {
"text": "large vocabulary speech recognition",
"start": 43,
"end": 78
}
}
]
}
} | [] |
First , images are partitioned into regions using one-class classification and patch-based clustering algorithms where one-class classifiers model the regions with relatively uniform color and texture properties , and clustering of patches aims to detect structures in the remaining regions . | {
"relations": {
"conjunction": [
{
"head": {
"text": "patch-based clustering algorithms",
"start": 79,
"end": 112
},
"tail": {
"text": "one-class classification",
"start": 50,
"end": 74
}
}
],
"used fo... | [] |
Next , the resulting regions are clustered to obtain a codebook of region types , and two models are constructed for scene representation : a '' bag of individual regions '' representation where each region is regarded separately , and a '' bag of region pairs '' representation where regions with particular spatial rel... | {
"relations": {
"used for": [
{
"head": {
"text": "models",
"start": 90,
"end": 96
},
"tail": {
"text": "scene representation",
"start": 117,
"end": 137
}
}
]
}
} | [] |
Given these representations , scene classification is done using Bayesian classifiers . | {
"relations": {
"used for": [
{
"head": {
"text": "Bayesian classifiers",
"start": 65,
"end": 85
},
"tail": {
"text": "scene classification",
"start": 30,
"end": 50
}
}
]
}
} | [] |
Experiments on the LabelMe data set showed that the proposed models significantly out-perform a baseline global feature-based approach . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "LabelMe data set",
"start": 19,
"end": 35
},
"tail": {
"text": "models",
"start": 61,
"end": 67
}
},
{
"head": {
"text": "LabelMe ... | [] |
The model is designed for use in error correction , with a focus on post-processing the output of black-box OCR systems in order to make it more useful for NLP tasks . | {
"relations": {
"used for": [
{
"head": {
"text": "model",
"start": 4,
"end": 9
},
"tail": {
"text": "error correction",
"start": 33,
"end": 49
}
},
{
"head": {
"text": "model",
... | [] |
We present an implementation of the model based on finite-state models , demonstrate the model 's ability to significantly reduce character and word error rate , and provide evaluation results involving automatic extraction of translation lexicons from printed text . | {
"relations": {
"used for": [
{
"head": {
"text": "finite-state models",
"start": 51,
"end": 70
},
"tail": {
"text": "model",
"start": 36,
"end": 41
}
},
{
"head": {
"text": "printed te... | [] |
We present a framework for word alignment based on log-linear models . | {
"relations": {
"used for": [
{
"head": {
"text": "framework",
"start": 13,
"end": 22
},
"tail": {
"text": "word alignment",
"start": 27,
"end": 41
}
},
{
"head": {
"text": "log-linear ... | [] |
All knowledge sources are treated as feature functions , which depend on the source langauge sentence , the target language sentence and possible additional variables . | {
"relations": {
"used for": [
{
"head": {
"text": "knowledge sources",
"start": 4,
"end": 21
},
"tail": {
"text": "feature functions",
"start": 37,
"end": 54
}
}
]
}
} | [] |
Log-linear models allow statistical alignment models to be easily extended by incorporating syntactic information . | {
"relations": {
"used for": [
{
"head": {
"text": "Log-linear models",
"start": 0,
"end": 17
},
"tail": {
"text": "statistical alignment models",
"start": 24,
"end": 52
}
},
{
"head": {
... | [] |
In this paper , we use IBM Model 3 alignment probabilities , POS correspondence , and bilingual dictionary coverage as features . | {
"relations": {
"conjunction": [
{
"head": {
"text": "IBM Model 3 alignment probabilities",
"start": 23,
"end": 58
},
"tail": {
"text": "POS correspondence",
"start": 61,
"end": 79
}
},
{
"head":... | [] |
Our experiments show that log-linear models significantly outperform IBM translation models . | {
"relations": {
"compare": [
{
"head": {
"text": "log-linear models",
"start": 26,
"end": 43
},
"tail": {
"text": "IBM translation models",
"start": 69,
"end": 91
}
}
]
}
} | [] |
Hough voting in a geometric transformation space allows us to realize spatial verification , but remains sensitive to feature detection errors because of the inflexible quan-tization of single feature correspondences . | {
"relations": {
"used for": [
{
"head": {
"text": "Hough voting",
"start": 0,
"end": 12
},
"tail": {
"text": "spatial verification",
"start": 70,
"end": 90
}
}
],
"feature of": [
{
"head"... | [] |
To handle this problem , we propose a new method , called adaptive dither voting , for robust spatial verification . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 42,
"end": 48
},
"tail": {
"text": "robust spatial verification",
"start": 87,
"end": 114
}
}
]
}
} | [] |
For each correspondence , instead of hard-mapping it to a single transformation , the method augments its description by using multiple dithered transformations that are deterministically generated by the other correspondences . | {
"relations": {
"used for": [
{
"head": {
"text": "multiple dithered transformations",
"start": 127,
"end": 160
},
"tail": {
"text": "method",
"start": 86,
"end": 92
}
}
]
}
} | [] |
We also propose exploiting the non-uniformity of a Hough histogram as the spatial similarity to handle multiple matching surfaces . | {
"relations": {
"feature of": [
{
"head": {
"text": "non-uniformity",
"start": 31,
"end": 45
},
"tail": {
"text": "Hough histogram",
"start": 51,
"end": 66
}
}
],
"used for": [
{
"head": ... | [] |
The method outperforms its state-of-the-art counterparts in both accuracy and scalability , especially when it comes to the retrieval of small , rotated objects . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 4,
"end": 10
},
"tail": {
"text": "retrieval of small , rotated objects",
"start": 124,
"end": 160
}
}
],
"compare": [
{
... | [] |
We propose a novel technique called bispectral photo-metric stereo that makes effective use of fluorescence for shape reconstruction . | {
"relations": {
"used for": [
{
"head": {
"text": "bispectral photo-metric stereo",
"start": 36,
"end": 66
},
"tail": {
"text": "shape reconstruction",
"start": 112,
"end": 132
}
},
{
"head": {
... | [] |
Due to the complexity of its emission process , fluo-rescence tends to be excluded from most algorithms in computer vision and image processing . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "complexity",
"start": 11,
"end": 21
},
"tail": {
"text": "emission process",
"start": 29,
"end": 45
}
}
],
"used for": [
{
"head": {... | [] |
Moreover , fluorescence 's wavelength-shifting property enables us to estimate the shape of an object by applying photomet-ric stereo to emission-only images without suffering from specular reflection . | {
"relations": {
"used for": [
{
"head": {
"text": "fluorescence 's wavelength-shifting property",
"start": 11,
"end": 55
},
"tail": {
"text": "shape",
"start": 83,
"end": 88
}
},
{
"head": {
... | [] |
This is the significant advantage of the fluorescence-based method over previous methods based on reflection . | {
"relations": {
"compare": [
{
"head": {
"text": "methods",
"start": 81,
"end": 88
},
"tail": {
"text": "fluorescence-based method",
"start": 41,
"end": 66
}
}
]
}
} | [] |
In this paper , we present an approach for learning a visual representation from the raw spatiotemporal signals in videos . | {
"relations": {
"used for": [
{
"head": {
"text": "approach",
"start": 30,
"end": 38
},
"tail": {
"text": "visual representation",
"start": 54,
"end": 75
}
},
{
"head": {
"text": "raw s... | [] |
We formulate our method as an unsupervised sequential verification task , i.e. , we determine whether a sequence of frames from a video is in the correct temporal order . | {
"relations": {
"used for": [
{
"head": {
"text": "unsupervised sequential verification task",
"start": 30,
"end": 71
},
"tail": {
"text": "method",
"start": 17,
"end": 23
}
}
]
}
} | [] |
With this simple task and no semantic labels , we learn a powerful visual representation using a Convolutional Neural Network -LRB- CNN -RRB- . | {
"relations": {
"used for": [
{
"head": {
"text": "task",
"start": 17,
"end": 21
},
"tail": {
"text": "visual representation",
"start": 67,
"end": 88
}
},
{
"head": {
"text": "Convoluti... | [] |
The representation contains complementary information to that learned from supervised image datasets like ImageNet . | {
"relations": {
"part of": [
{
"head": {
"text": "complementary information",
"start": 28,
"end": 53
},
"tail": {
"text": "representation",
"start": 4,
"end": 18
}
}
],
"used for": [
{
"h... | [] |
Qualitative results show that our method captures information that is temporally varying , such as human pose . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 34,
"end": 40
},
"tail": {
"text": "human pose",
"start": 99,
"end": 109
}
}
]
}
} | [] |
When used as pre-training for action recognition , our method gives significant gains over learning without external data on benchmark datasets like UCF101 and HMDB51 . | {
"relations": {
"used for": [
{
"head": {
"text": "pre-training",
"start": 13,
"end": 25
},
"tail": {
"text": "action recognition",
"start": 30,
"end": 48
}
},
{
"head": {
"text": "our ... | [] |
To demonstrate its sensitivity to human pose , we show results for pose estimation on the FLIC and MPII datasets that are competitive , or better than approaches using significantly more supervision . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "FLIC and MPII datasets",
"start": 90,
"end": 112
},
"tail": {
"text": "pose estimation",
"start": 67,
"end": 82
}
}
],
"used for": [
{
... | [] |
Our method can be combined with supervised representations to provide an additional boost in accuracy . | {
"relations": {
"conjunction": [
{
"head": {
"text": "supervised representations",
"start": 32,
"end": 58
},
"tail": {
"text": "Our method",
"start": 0,
"end": 10
}
}
],
"evaluate for": [
{
... | [] |
`` To explain complex phenomena , an explanation system must be able to select information from a formal representation of domain knowledge , organize the selected information into multisentential discourse plans , and realize the discourse plans in text . | {
"relations": {
"used for": [
{
"head": {
"text": "explanation system",
"start": 37,
"end": 55
},
"tail": {
"text": "discourse plans",
"start": 197,
"end": 212
}
}
]
}
} | [] |
This paper reports on a seven-year effort to empirically study explanation generation from semantically rich , large-scale knowledge bases . | {
"relations": {
"used for": [
{
"head": {
"text": "semantically rich , large-scale knowledge bases",
"start": 91,
"end": 138
},
"tail": {
"text": "explanation generation",
"start": 63,
"end": 85
}
}
]
}
... | [] |
In particular , it describes a robust explanation system that constructs multisentential and multi-paragraph explanations from the a large-scale knowledge base in the domain of botanical anatomy , physiology , and development . | {
"relations": {
"used for": [
{
"head": {
"text": "robust explanation system",
"start": 31,
"end": 56
},
"tail": {
"text": "multisentential and multi-paragraph explanations",
"start": 73,
"end": 121
}
},
... | [] |
We introduce the evaluation methodology and describe how performance was assessed with this methodology in the most extensive empirical evaluation conducted on an explanation system . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "methodology",
"start": 28,
"end": 39
},
"tail": {
"text": "explanation system",
"start": 163,
"end": 181
}
}
]
}
} | [] |
We present an operable definition of focus which is argued to be of a cognito-pragmatic nature and explore how it is determined in discourse in a formalized manner . | {
"relations": {
"feature of": [
{
"head": {
"text": "cognito-pragmatic nature",
"start": 70,
"end": 94
},
"tail": {
"text": "operable definition",
"start": 14,
"end": 33
}
}
]
}
} | [] |
For this purpose , a file card model of discourse model and knowledge store is introduced enabling the decomposition and formal representation of its determination process as a programmable algorithm -LRB- FDA -RRB- . | {
"relations": {
"hyponym of": [
{
"head": {
"text": "FDA",
"start": 206,
"end": 209
},
"tail": {
"text": "programmable algorithm",
"start": 177,
"end": 199
}
}
]
}
} | [] |
Interdisciplinary evidence from social and cognitive psychology is cited and the prospect of the integration of focus via FDA as a discourse-level construct into speech synthesis systems , in particular , concept-to-speech systems , is also briefly discussed . | {
"relations": {
"used for": [
{
"head": {
"text": "FDA",
"start": 122,
"end": 125
},
"tail": {
"text": "discourse-level construct",
"start": 131,
"end": 156
}
}
],
"part of": [
{
"head": ... | [] |
Inference in these models involves solving a combinatorial optimization problem , with methods such as graph cuts , belief propagation . | {
"relations": {
"used for": [
{
"head": {
"text": "Inference",
"start": 0,
"end": 9
},
"tail": {
"text": "models",
"start": 19,
"end": 25
}
},
{
"head": {
"text": "methods",
"... | [] |
To overcome this , state-of-the-art structured learning methods frame the problem as one of large margin estimation . | {
"relations": {
"used for": [
{
"head": {
"text": "structured learning methods",
"start": 36,
"end": 63
},
"tail": {
"text": "problem",
"start": 74,
"end": 81
}
},
{
"head": {
"text": "... | [] |
Iterative solutions have been proposed to solve the resulting convex optimization problem . | {
"relations": {
"used for": [
{
"head": {
"text": "Iterative solutions",
"start": 0,
"end": 19
},
"tail": {
"text": "convex optimization problem",
"start": 62,
"end": 89
}
}
]
}
} | [] |
We show how the resulting optimization problem can be reduced to an equivalent convex problem with a small number of constraints , and solve it using an efficient scheme . | {
"relations": {
"used for": [
{
"head": {
"text": "convex problem",
"start": 79,
"end": 93
},
"tail": {
"text": "optimization problem",
"start": 26,
"end": 46
}
}
]
}
} | [] |
Interpreting metaphors is an integral and inescapable process in human understanding of natural language . | {
"relations": {
"hyponym of": [
{
"head": {
"text": "Interpreting metaphors",
"start": 0,
"end": 22
},
"tail": {
"text": "human understanding of natural language",
"start": 65,
"end": 104
}
}
]
}
} | [] |
This paper discusses a method of analyzing metaphors based on the existence of a small number of generalized metaphor mappings . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 23,
"end": 29
},
"tail": {
"text": "analyzing metaphors",
"start": 33,
"end": 52
}
},
{
"head": {
"text": "generaliz... | [] |
Each generalized metaphor contains a recognition network , a basic mapping , additional transfer mappings , and an implicit intention component . | {
"relations": {
"part of": [
{
"head": {
"text": "recognition network",
"start": 37,
"end": 56
},
"tail": {
"text": "generalized metaphor",
"start": 5,
"end": 25
}
},
{
"head": {
"text"... | [] |
It is argued that the method reduces metaphor interpretation from a reconstruction to a recognition task . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 22,
"end": 28
},
"tail": {
"text": "metaphor interpretation",
"start": 37,
"end": 60
}
},
{
"head": {
"text": "recog... | [] |
This study presents a method to automatically acquire paraphrases using bilingual corpora , which utilizes the bilingual dependency relations obtained by projecting a monolingual dependency parse onto the other language sentence based on statistical alignment techniques . | {
"relations": {
"used for": [
{
"head": {
"text": "bilingual corpora",
"start": 72,
"end": 89
},
"tail": {
"text": "method to automatically acquire paraphrases",
"start": 22,
"end": 65
}
},
{
"he... | [] |
Since the paraphrasing method is capable of clearly disambiguating the sense of an original phrase using the bilingual context of dependency relation , it would be possible to obtain interchangeable paraphrases under a given context . | {
"relations": {
"used for": [
{
"head": {
"text": "bilingual context of dependency relation",
"start": 109,
"end": 149
},
"tail": {
"text": "paraphrasing method",
"start": 10,
"end": 29
}
}
]
}
} | [] |
Also , we provide an advanced method to acquire generalized translation knowledge using the extracted paraphrases . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 30,
"end": 36
},
"tail": {
"text": "generalized translation knowledge",
"start": 48,
"end": 81
}
},
{
"head": {
"tex... | [] |
We applied the method to acquire the generalized translation knowledge for Korean-English translation . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 15,
"end": 21
},
"tail": {
"text": "generalized translation knowledge",
"start": 37,
"end": 70
}
},
{
"head": {
"tex... | [] |
Through experiments with parallel corpora of a Korean and English language pairs , we show that our paraphrasing method effectively extracts paraphrases with high precision , 94.3 % and 84.6 % respectively for Korean and English , and the translation knowledge extracted from the bilingual corpora could be generalized s... | {
"relations": {
"used for": [
{
"head": {
"text": "paraphrasing method",
"start": 100,
"end": 119
},
"tail": {
"text": "paraphrases",
"start": 141,
"end": 152
}
},
{
"head": {
"text": "... | [] |
We provide a logical definition of Minimalist grammars , that are Stabler 's formalization of Chomsky 's minimalist program . | {
"relations": {
"hyponym of": [
{
"head": {
"text": "Stabler 's formalization of Chomsky 's minimalist program",
"start": 66,
"end": 123
},
"tail": {
"text": "logical definition of Minimalist grammars",
"start": 13,
"end": ... | [] |
Our logical definition leads to a neat relation to categorial grammar , -LRB- yielding a treatment of Montague semantics -RRB- , a parsing-as-deduction in a resource sensitive logic , and a learning algorithm from structured data -LRB- based on a typing-algorithm and type-unification -RRB- . | {
"relations": {
"used for": [
{
"head": {
"text": "logical definition",
"start": 4,
"end": 22
},
"tail": {
"text": "Montague semantics",
"start": 102,
"end": 120
}
},
{
"head": {
"text"... | [] |
There are several approaches that model information extraction as a token classification task , using various tagging strategies to combine multiple tokens . | {
"relations": {
"used for": [
{
"head": {
"text": "approaches",
"start": 18,
"end": 28
},
"tail": {
"text": "information extraction",
"start": 40,
"end": 62
}
},
{
"head": {
"text": "ta... | [] |
We also introduce a new strategy , called Begin/After tagging or BIA , and show that it is competitive to the best other strategies . | {
"relations": {
"compare": [
{
"head": {
"text": "it",
"start": 85,
"end": 87
},
"tail": {
"text": "strategies",
"start": 121,
"end": 131
}
}
]
}
} | [] |
The objective is a generic system of tools , including a core English lexicon , grammar , and concept representations , for building natural language processing -LRB- NLP -RRB- systems for text understanding . | {
"relations": {
"used for": [
{
"head": {
"text": "system",
"start": 27,
"end": 33
},
"tail": {
"text": "natural language processing -LRB- NLP -RRB- systems",
"start": 133,
"end": 184
}
},
{
"hea... | [] |
Systems built with PAKTUS are intended to generate input to knowledge based systems ordata base systems . | {
"relations": {
"used for": [
{
"head": {
"text": "PAKTUS",
"start": 19,
"end": 25
},
"tail": {
"text": "knowledge based systems",
"start": 60,
"end": 83
}
}
]
}
} | [] |
Input to the NLP system is typically derived from an existing electronic message stream , such as a news wire . | {
"relations": {
"used for": [
{
"head": {
"text": "electronic message stream",
"start": 62,
"end": 87
},
"tail": {
"text": "NLP system",
"start": 13,
"end": 23
}
}
],
"hyponym of": [
{
"h... | [] |
PAKTUS supports the adaptation of the generic core to a variety of domains : JINTACCS messages , RAINFORM messages , news reports about a specific type of event , such as financial transfers or terrorist acts , etc. , by acquiring sublanguage and domain-specific grammar , words , conceptual mappings , and discourse pat... | {
"relations": {
"used for": [
{
"head": {
"text": "PAKTUS",
"start": 0,
"end": 6
},
"tail": {
"text": "JINTACCS messages",
"start": 77,
"end": 94
}
},
{
"head": {
"text": "PAKTUS",
... | [] |
In this paper the LIMSI recognizer which was evaluated in the ARPA NOV93 CSR test is described , and experimental results on the WSJ and BREF corpora under closely matched conditions are reported . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "ARPA NOV93 CSR test",
"start": 62,
"end": 81
},
"tail": {
"text": "LIMSI recognizer",
"start": 18,
"end": 34
}
},
{
"head": {
"tex... | [] |
For both corpora word recognition experiments were carried out with vocabularies containing up to 20k words . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "corpora",
"start": 9,
"end": 16
},
"tail": {
"text": "word recognition",
"start": 17,
"end": 33
}
}
]
}
} | [] |
The recognizer makes use of continuous density HMM with Gaussian mixture for acoustic modeling and n-gram statistics estimated on the newspaper texts for language modeling . | {
"relations": {
"used for": [
{
"head": {
"text": "continuous density HMM",
"start": 28,
"end": 50
},
"tail": {
"text": "recognizer",
"start": 4,
"end": 14
}
},
{
"head": {
"text": "con... | [] |
The recognizer uses a time-synchronous graph-search strategy which is shown to still be viable with a 20k-word vocabulary when used with bigram back-off language models . | {
"relations": {
"used for": [
{
"head": {
"text": "time-synchronous graph-search strategy",
"start": 22,
"end": 60
},
"tail": {
"text": "recognizer",
"start": 4,
"end": 14
}
},
{
"head": {
... | [] |
A second forward pass , which makes use of a word graph generated with the bigram , incorporates a trigram language model . | {
"relations": {
"used for": [
{
"head": {
"text": "bigram",
"start": 75,
"end": 81
},
"tail": {
"text": "word graph",
"start": 45,
"end": 55
}
}
],
"conjunction": [
{
"head": {
... | [] |
Acoustic modeling uses cepstrum-based features , context-dependent phone models -LRB- intra and interword -RRB- , phone duration models , and sex-dependent models . | {
"relations": {
"used for": [
{
"head": {
"text": "cepstrum-based features",
"start": 23,
"end": 46
},
"tail": {
"text": "Acoustic modeling",
"start": 0,
"end": 17
}
},
{
"head": {
"tex... | [] |
The co-occurrence pattern , a combination of binary or local features , is more discriminative than individual features and has shown its advantages in object , scene , and action recognition . | {
"relations": {
"used for": [
{
"head": {
"text": "co-occurrence pattern",
"start": 4,
"end": 25
},
"tail": {
"text": "object , scene , and action recognition",
"start": 152,
"end": 191
}
}
],
"part of... | [] |
Then we propose a novel data mining method to efficiently discover the optimal co-occurrence pattern with minimum empirical error , despite the noisy training dataset . | {
"relations": {
"used for": [
{
"head": {
"text": "data mining method",
"start": 24,
"end": 42
},
"tail": {
"text": "optimal co-occurrence pattern",
"start": 71,
"end": 100
}
},
{
"head": {
... | [] |
This mining procedure of AND and OR patterns is readily integrated to boosting , which improves the generalization ability over the conventional boosting decision trees and boosting decision stumps . | {
"relations": {
"used for": [
{
"head": {
"text": "mining procedure",
"start": 5,
"end": 21
},
"tail": {
"text": "AND and OR patterns",
"start": 25,
"end": 44
}
}
],
"part of": [
{
"head"... | [] |
Our versatile experiments on object , scene , and action cat-egorization validate the advantages of the discovered dis-criminative co-occurrence patterns . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "object , scene , and action cat-egorization",
"start": 29,
"end": 72
},
"tail": {
"text": "dis-criminative co-occurrence patterns",
"start": 115,
"end": 153
}
... | [] |
Empirical experience and observations have shown us when powerful and highly tunable classifiers such as maximum entropy classifiers , boosting and SVMs are applied to language processing tasks , it is possible to achieve high accuracies , but eventually their performances all tend to plateau out at around the same poi... | {
"relations": {
"used for": [
{
"head": {
"text": "classifiers",
"start": 85,
"end": 96
},
"tail": {
"text": "language processing tasks",
"start": 168,
"end": 193
}
}
],
"hyponym of": [
{
... | [] |
In recent work , we introduced N-fold Templated Piped Correction , or NTPC -LRB- `` nitpick '' -RRB- , an intriguing error corrector that is designed to work in these extreme operating conditions . | {
"relations": {
"hyponym of": [
{
"head": {
"text": "N-fold Templated Piped Correction , or NTPC -LRB- `` nitpick '' -RRB-",
"start": 31,
"end": 100
},
"tail": {
"text": "error corrector",
"start": 117,
"end": 132
}... | [] |
Despite its simplicity , it consistently and robustly improves the accuracy of existing highly accurate base models . | {
"relations": {
"compare": [
{
"head": {
"text": "it",
"start": 4,
"end": 6
},
"tail": {
"text": "base models",
"start": 104,
"end": 115
}
}
]
}
} | [] |
Focused interaction of this kind is facilitated by a construction-specific approach to flexible parsing , with specialized parsing techniques for each type of construction , and specialized ambiguity representations for each type of ambiguity that a particular construction can give rise to . | {
"relations": {
"used for": [
{
"head": {
"text": "construction-specific approach",
"start": 53,
"end": 83
},
"tail": {
"text": "flexible parsing",
"start": 87,
"end": 103
}
},
{
"head": {
... | [] |
A construction-specific approach also aids in task-specific language development by allowing a language definition that is natural in terms of the task domain to be interpreted directly without compilation into a uniform grammar formalism , thus greatly speeding the testing of changes to the language definition . | {
"relations": {
"used for": [
{
"head": {
"text": "construction-specific approach",
"start": 2,
"end": 32
},
"tail": {
"text": "task-specific language development",
"start": 46,
"end": 80
}
}
]
}
} | [] |
A proposal to deal with French tenses in the framework of Discourse Representation Theory is presented , as it has been implemented for a fragment at the IMS . | {
"relations": {
"used for": [
{
"head": {
"text": "Discourse Representation Theory",
"start": 58,
"end": 89
},
"tail": {
"text": "French tenses",
"start": 24,
"end": 37
}
},
{
"head": {
... | [] |
It is based on the theory of tenses of H. Kamp and Ch . | {
"relations": {
"used for": [
{
"head": {
"text": "theory of tenses",
"start": 19,
"end": 35
},
"tail": {
"text": "It",
"start": 0,
"end": 2
}
}
]
}
} | [] |
Instead of using operators to express the meaning of the tenses the Reichenbachian point of view is adopted and refined such that the impact of the tenses with respect to the meaning of the text is understood as contribution to the integration of the events of a sentence in the event structure of the preceeding text . | {
"relations": {
"used for": [
{
"head": {
"text": "operators",
"start": 17,
"end": 26
},
"tail": {
"text": "meaning of the tenses",
"start": 42,
"end": 63
}
}
]
}
} | [] |
Thereby a system of relevant times provided by the preceeding text and by the temporal adverbials of the sentence being processed is used . | {
"relations": {
"used for": [
{
"head": {
"text": "preceeding text",
"start": 51,
"end": 66
},
"tail": {
"text": "system of relevant times",
"start": 10,
"end": 34
}
},
{
"head": {
"tex... | [] |
This system consists of one or more reference times and temporal perspective times , the speech time and the location time . | {
"relations": {
"part of": [
{
"head": {
"text": "reference times",
"start": 36,
"end": 51
},
"tail": {
"text": "system",
"start": 5,
"end": 11
}
},
{
"head": {
"text": "temporal perspe... | [] |
In opposition to the approach of Kamp and Rohrer the exact meaning of the tenses is fixed by the resolution component and not in the process of syntactic analysis . | {
"relations": {
"used for": [
{
"head": {
"text": "resolution component",
"start": 97,
"end": 117
},
"tail": {
"text": "meaning of the tenses",
"start": 59,
"end": 80
}
}
],
"compare": [
{
... | [] |
The work presented in this paper is the first step in a project which aims to cluster and summarise electronic discussions in the context of help-desk applications . | {
"relations": {
"part of": [
{
"head": {
"text": "electronic discussions",
"start": 100,
"end": 122
},
"tail": {
"text": "help-desk applications",
"start": 141,
"end": 163
}
}
]
}
} | [] |
In this paper , we identify features of electronic discussions that influence the clustering process , and offer a filtering mechanism that removes undesirable influences . | {
"relations": {
"feature of": [
{
"head": {
"text": "features",
"start": 28,
"end": 36
},
"tail": {
"text": "electronic discussions",
"start": 40,
"end": 62
}
}
]
}
} | [] |
We tested the clustering and filtering processes on electronic newsgroup discussions , and evaluated their performance by means of two experiments : coarse-level clustering simple information retrieval . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "electronic newsgroup discussions",
"start": 52,
"end": 84
},
"tail": {
"text": "clustering and filtering processes",
"start": 14,
"end": 48
}
},
{
... | [] |
The paper presents a method for word sense disambiguation based on parallel corpora . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 21,
"end": 27
},
"tail": {
"text": "word sense disambiguation",
"start": 32,
"end": 57
}
},
{
"head": {
"text": "par... | [] |
The method exploits recent advances in word alignment and word clustering based on automatic extraction of translation equivalents and being supported by available aligned wordnets for the languages in the corpus . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 4,
"end": 10
},
"tail": {
"text": "word alignment",
"start": 39,
"end": 53
}
},
{
"head": {
"text": "method",
... | [] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.