input stringlengths 37 565 | output dict | schema listlengths 0 0 |
|---|---|---|
This method is also capable of handling unknown words , which is important in practical systems . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 5,
"end": 11
},
"tail": {
"text": "unknown words",
"start": 40,
"end": 53
}
}
]
}
} | [] |
This paper shows that it is very often possible to identify the source language of medium-length speeches in the EUROPARL corpus on the basis of frequency counts of word n-grams -LRB- 87.2 % -96.7 % accuracy depending on classification method -RRB- . | {
"relations": {
"part of": [
{
"head": {
"text": "medium-length speeches",
"start": 83,
"end": 105
},
"tail": {
"text": "EUROPARL corpus",
"start": 113,
"end": 128
}
}
],
"evaluate for": [
{
... | [] |
We investigated whether automatic phonetic transcriptions -LRB- APTs -RRB- can replace manually verified phonetic transcriptions -LRB- MPTs -RRB- in a large corpus-based study on pronunciation variation . | {
"relations": {
"compare": [
{
"head": {
"text": "automatic phonetic transcriptions -LRB- APTs -RRB-",
"start": 24,
"end": 74
},
"tail": {
"text": "manually verified phonetic transcriptions",
"start": 87,
"end": 128
... | [] |
We trained classifiers on the speech processes extracted from the alignments of an APT and an MPT with a canonical transcription . | {
"relations": {
"used for": [
{
"head": {
"text": "speech processes",
"start": 30,
"end": 46
},
"tail": {
"text": "classifiers",
"start": 11,
"end": 22
}
},
{
"head": {
"text": "alignme... | [] |
We tested whether the classifiers were equally good at verifying whether unknown transcriptions represent read speech or telephone dialogues , and whether the same speech processes were identified to distinguish between transcriptions of the two situational settings . | {
"relations": {
"used for": [
{
"head": {
"text": "classifiers",
"start": 22,
"end": 33
},
"tail": {
"text": "unknown transcriptions",
"start": 73,
"end": 95
}
},
{
"head": {
"text": "u... | [] |
Our results not only show that similar distinguishing speech processes were identified ; our APT-based classifier yielded better classification accuracy than the MPT-based classifier whilst using fewer classification features . | {
"relations": {
"compare": [
{
"head": {
"text": "APT-based classifier",
"start": 93,
"end": 113
},
"tail": {
"text": "MPT-based classifier",
"start": 162,
"end": 182
}
}
],
"evaluate for": [
{
... | [] |
Machine reading is a relatively new field that features computer programs designed to read flowing text and extract fact assertions expressed by the narrative content . | {
"relations": {
"used for": [
{
"head": {
"text": "computer programs",
"start": 56,
"end": 73
},
"tail": {
"text": "flowing text",
"start": 91,
"end": 103
}
},
{
"head": {
"text": "comp... | [] |
This task involves two core technologies : natural language processing -LRB- NLP -RRB- and information extraction -LRB- IE -RRB- . | {
"relations": {
"part of": [
{
"head": {
"text": "natural language processing -LRB- NLP -RRB-",
"start": 43,
"end": 86
},
"tail": {
"text": "task",
"start": 5,
"end": 9
}
},
{
"head": {
... | [] |
In this paper we describe a machine reading system that we have developed within a cognitive architecture . | {
"relations": {
"feature of": [
{
"head": {
"text": "cognitive architecture",
"start": 83,
"end": 105
},
"tail": {
"text": "machine reading system",
"start": 28,
"end": 50
}
}
]
}
} | [] |
We show how we have integrated into the framework several levels of knowledge for a particular domain , ideas from cognitive semantics and construction grammar , plus tools from prior NLP and IE research . | {
"relations": {
"conjunction": [
{
"head": {
"text": "cognitive semantics",
"start": 115,
"end": 134
},
"tail": {
"text": "construction grammar",
"start": 139,
"end": 159
}
},
{
"head": {
... | [] |
The result is a system that is capable of reading and interpreting complex and fairly idiosyncratic texts in the family history domain . | {
"relations": {
"used for": [
{
"head": {
"text": "system",
"start": 16,
"end": 22
},
"tail": {
"text": "idiosyncratic texts",
"start": 86,
"end": 105
}
}
],
"feature of": [
{
"head": {
... | [] |
We present two methods for capturing nonstationary chaos , then present a few examples including biological signals , ocean waves and traffic flow . | {
"relations": {
"used for": [
{
"head": {
"text": "methods",
"start": 15,
"end": 22
},
"tail": {
"text": "nonstationary chaos",
"start": 37,
"end": 56
}
}
],
"hyponym of": [
{
"head": {
... | [] |
This paper presents a formal analysis for a large class of words called alternative markers , which includes other -LRB- than -RRB- , such -LRB- as -RRB- , and besides . | {
"relations": {
"used for": [
{
"head": {
"text": "formal analysis",
"start": 22,
"end": 37
},
"tail": {
"text": "alternative markers",
"start": 72,
"end": 91
}
}
]
}
} | [] |
These words appear frequently enough in dialog to warrant serious attention , yet present natural language search engines perform poorly on queries containing them . | {
"relations": {
"part of": [
{
"head": {
"text": "words",
"start": 6,
"end": 11
},
"tail": {
"text": "dialog",
"start": 40,
"end": 46
}
}
]
}
} | [] |
I show that the performance of a search engine can be improved dramatically by incorporating an approximation of the formal analysis that is compatible with the search engine 's operational semantics . | {
"relations": {
"part of": [
{
"head": {
"text": "approximation of the formal analysis",
"start": 96,
"end": 132
},
"tail": {
"text": "search engine",
"start": 33,
"end": 46
}
},
{
"head": {
... | [] |
The value of this approach is that as the operational semantics of natural language applications improve , even larger improvements are possible . | {
"relations": {
"part of": [
{
"head": {
"text": "operational semantics",
"start": 42,
"end": 63
},
"tail": {
"text": "natural language applications",
"start": 67,
"end": 96
}
}
]
}
} | [] |
We find that simple interpolation methods , like log-linear and linear interpolation , improve the performance but fall short of the performance of an oracle . | {
"relations": {
"hyponym of": [
{
"head": {
"text": "log-linear and linear interpolation",
"start": 49,
"end": 84
},
"tail": {
"text": "interpolation methods",
"start": 20,
"end": 41
}
}
]
}
} | [] |
Actually , the oracle acts like a dynamic combiner with hard decisions using the reference . | {
"relations": {
"feature of": [
{
"head": {
"text": "hard decisions",
"start": 56,
"end": 70
},
"tail": {
"text": "dynamic combiner",
"start": 34,
"end": 50
}
}
]
}
} | [] |
We suggest a method that mimics the behavior of the oracle using a neural network or a decision tree . | {
"relations": {
"used for": [
{
"head": {
"text": "neural network",
"start": 67,
"end": 81
},
"tail": {
"text": "method",
"start": 13,
"end": 19
}
},
{
"head": {
"text": "decision tree"... | [] |
The method amounts to tagging LMs with confidence measures and picking the best hypothesis corresponding to the LM with the best confidence . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 4,
"end": 10
},
"tail": {
"text": "LMs",
"start": 30,
"end": 33
}
},
{
"head": {
"text": "confidence measures",
... | [] |
We describe a new method for the representation of NLP structures within reranking approaches . | {
"relations": {
"used for": [
{
"head": {
"text": "method",
"start": 18,
"end": 24
},
"tail": {
"text": "NLP structures",
"start": 51,
"end": 65
}
}
],
"feature of": [
{
"head": {
... | [] |
We make use of a conditional log-linear model , with hidden variables representing the assignment of lexical items to word clusters or word senses . | {
"relations": {
"used for": [
{
"head": {
"text": "hidden variables",
"start": 53,
"end": 69
},
"tail": {
"text": "conditional log-linear model",
"start": 17,
"end": 45
}
}
],
"conjunction": [
{
... | [] |
The model learns to automatically make these assignments based on a discriminative training criterion . | {
"relations": {
"used for": [
{
"head": {
"text": "discriminative training criterion",
"start": 68,
"end": 101
},
"tail": {
"text": "model",
"start": 4,
"end": 9
}
}
]
}
} | [] |
Training and decoding with the model requires summing over an exponential number of hidden-variable assignments : the required summations can be computed efficiently and exactly using dynamic programming . | {
"relations": {
"used for": [
{
"head": {
"text": "dynamic programming",
"start": 184,
"end": 203
},
"tail": {
"text": "summations",
"start": 127,
"end": 137
}
}
]
}
} | [] |
As a case study , we apply the model to parse reranking . | {
"relations": {
"used for": [
{
"head": {
"text": "model",
"start": 31,
"end": 36
},
"tail": {
"text": "parse reranking",
"start": 40,
"end": 55
}
}
]
}
} | [] |
The model gives an F-measure improvement of ~ 1.25 % beyond the base parser , and an ~ 0.25 % improvement beyond Collins -LRB- 2000 -RRB- reranker . | {
"relations": {
"compare": [
{
"head": {
"text": "model",
"start": 4,
"end": 9
},
"tail": {
"text": "base parser",
"start": 64,
"end": 75
}
},
{
"head": {
"text": "base parser",
... | [] |
Although our experiments are focused on parsing , the techniques described generalize naturally to NLP structures other than parse trees . | {
"relations": {
"used for": [
{
"head": {
"text": "techniques",
"start": 54,
"end": 64
},
"tail": {
"text": "parsing",
"start": 40,
"end": 47
}
},
{
"head": {
"text": "techniques",
... | [] |
This paper presents an algorithm for learning the time-varying shape of a non-rigid 3D object from uncalibrated 2D tracking data . | {
"relations": {
"used for": [
{
"head": {
"text": "algorithm",
"start": 23,
"end": 32
},
"tail": {
"text": "learning the time-varying shape of a non-rigid 3D object",
"start": 37,
"end": 93
}
}
]
}
} | [] |
We constrain the problem by assuming that the object shape at each time instant is drawn from a Gaussian distribution . | {
"relations": {
"used for": [
{
"head": {
"text": "Gaussian distribution",
"start": 96,
"end": 117
},
"tail": {
"text": "object shape",
"start": 46,
"end": 58
}
}
]
}
} | [] |
Based on this assumption , the algorithm simultaneously estimates 3D shape and motion for each time frame , learns the parameters of the Gaussian , and robustly fills-in missing data points . | {
"relations": {
"used for": [
{
"head": {
"text": "algorithm",
"start": 31,
"end": 40
},
"tail": {
"text": "3D shape and motion",
"start": 66,
"end": 85
}
}
]
}
} | [] |
We then extend the algorithm to model temporal smoothness in object shape , thus allowing it to handle severe cases of missing data . | {
"relations": {
"used for": [
{
"head": {
"text": "algorithm",
"start": 19,
"end": 28
},
"tail": {
"text": "temporal smoothness in object shape",
"start": 38,
"end": 73
}
},
{
"head": {
... | [] |
Automatic summarization and information extraction are two important Internet services . | {
"relations": {
"conjunction": [
{
"head": {
"text": "Automatic summarization",
"start": 0,
"end": 23
},
"tail": {
"text": "information extraction",
"start": 28,
"end": 50
}
}
]
}
} | [] |
MUC and SUMMAC play their appropriate roles in the next generation Internet . | {
"relations": {
"conjunction": [
{
"head": {
"text": "MUC",
"start": 0,
"end": 3
},
"tail": {
"text": "SUMMAC",
"start": 8,
"end": 14
}
}
]
}
} | [] |
This paper focuses on the automatic summarization and proposes two different models to extract sentences for summary generation under two tasks initiated by SUMMAC-1 . | {
"relations": {
"used for": [
{
"head": {
"text": "models",
"start": 77,
"end": 83
},
"tail": {
"text": "summary generation",
"start": 109,
"end": 127
}
},
{
"head": {
"text": "models",... | [] |
For categorization task , positive feature vectors and negative feature vectors are used cooperatively to construct generic , indicative summaries . | {
"relations": {
"used for": [
{
"head": {
"text": "positive feature vectors",
"start": 26,
"end": 50
},
"tail": {
"text": "categorization task",
"start": 4,
"end": 23
}
},
{
"head": {
"... | [] |
For adhoc task , a text model based on relationship between nouns and verbs is used to filter out irrelevant discourse segment , to rank relevant sentences , and to generate the user-directed summaries . | {
"relations": {
"used for": [
{
"head": {
"text": "text model",
"start": 19,
"end": 29
},
"tail": {
"text": "adhoc task",
"start": 4,
"end": 14
}
},
{
"head": {
"text": "text model",
... | [] |
The result shows that the NormF of the best summary and that of the fixed summary for adhoc tasks are 0.456 and 0 . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "NormF",
"start": 26,
"end": 31
},
"tail": {
"text": "adhoc tasks",
"start": 86,
"end": 97
}
}
]
}
} | [] |
The NormF of the best summary and that of the fixed summary for categorization task are 0.4090 and 0.4023 . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "NormF",
"start": 4,
"end": 9
},
"tail": {
"text": "categorization task",
"start": 64,
"end": 83
}
}
]
}
} | [] |
Our system outperforms the average system in categorization task but does a common job in adhoc task . | {
"relations": {
"compare": [
{
"head": {
"text": "system",
"start": 4,
"end": 10
},
"tail": {
"text": "system",
"start": 4,
"end": 10
}
}
],
"evaluate for": [
{
"head": {
"text"... | [] |
In real-world action recognition problems , low-level features can not adequately characterize the rich spatial-temporal structures in action videos . | {
"relations": {
"feature of": [
{
"head": {
"text": "rich spatial-temporal structures",
"start": 99,
"end": 131
},
"tail": {
"text": "action videos",
"start": 135,
"end": 148
}
}
]
}
} | [] |
The second type is data-driven attributes , which are learned from data using dictionary learning methods . | {
"relations": {
"used for": [
{
"head": {
"text": "dictionary learning methods",
"start": 78,
"end": 105
},
"tail": {
"text": "data-driven attributes",
"start": 19,
"end": 41
}
}
]
}
} | [] |
We propose a discriminative and compact attribute-based representation by selecting a subset of discriminative attributes from a large attribute set . | {
"relations": {
"used for": [
{
"head": {
"text": "discriminative attributes",
"start": 96,
"end": 121
},
"tail": {
"text": "discriminative and compact attribute-based representation",
"start": 13,
"end": 70
}
... | [] |
Three attribute selection criteria are proposed and formulated as a submodular optimization problem . | {
"relations": {
"used for": [
{
"head": {
"text": "submodular optimization problem",
"start": 68,
"end": 99
},
"tail": {
"text": "attribute selection criteria",
"start": 6,
"end": 34
}
}
]
}
} | [] |
Experimental results on the Olympic Sports and UCF101 datasets demonstrate that the proposed attribute-based representation can significantly boost the performance of action recognition algorithms and outperform most recently proposed recognition approaches . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "Olympic Sports and UCF101 datasets",
"start": 28,
"end": 62
},
"tail": {
"text": "attribute-based representation",
"start": 93,
"end": 123
}
}
],
... | [] |
Landsbergen 's advocacy of analytical inverses for compositional syntax rules encourages the application of Definite Clause Grammar techniques to the construction of a parser returning Montague analysis trees . | {
"relations": {
"used for": [
{
"head": {
"text": "analytical inverses",
"start": 27,
"end": 46
},
"tail": {
"text": "compositional syntax rules",
"start": 51,
"end": 77
}
},
{
"head": {
... | [] |
A parser MDCC is presented which implements an augmented Friedman - Warren algorithm permitting post referencing * and interfaces with a language of intenslonal logic translator LILT so as to display the derivational history of corresponding reduced IL formulae . | {
"relations": {
"used for": [
{
"head": {
"text": "augmented Friedman - Warren algorithm",
"start": 47,
"end": 84
},
"tail": {
"text": "parser MDCC",
"start": 2,
"end": 13
}
},
{
"head": {
... | [] |
Some familiarity with Montague 's PTQ and the basic DCG mechanism is assumed . | {
"relations": {
"conjunction": [
{
"head": {
"text": "Montague 's PTQ",
"start": 22,
"end": 37
},
"tail": {
"text": "basic DCG mechanism",
"start": 46,
"end": 65
}
}
]
}
} | [] |
Stochastic attention-based models have been shown to improve computational efficiency at test time , but they remain difficult to train because of intractable posterior inference and high variance in the stochastic gradient estimates . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "computational efficiency",
"start": 61,
"end": 85
},
"tail": {
"text": "Stochastic attention-based models",
"start": 0,
"end": 33
}
}
],
"conjunct... | [] |
Borrowing techniques from the literature on training deep generative models , we present the Wake-Sleep Recurrent Attention Model , a method for training stochastic attention networks which improves posterior inference and which reduces the variability in the stochastic gradients . | {
"relations": {
"used for": [
{
"head": {
"text": "Borrowing techniques",
"start": 0,
"end": 20
},
"tail": {
"text": "deep generative models",
"start": 53,
"end": 75
}
},
{
"head": {
"t... | [] |
We show that our method can greatly speed up the training time for stochastic attention networks in the domains of image classification and caption generation . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "training time",
"start": 49,
"end": 62
},
"tail": {
"text": "method",
"start": 17,
"end": 23
}
},
{
"head": {
"text": "image class... | [] |
A new exemplar-based framework unifying image completion , texture synthesis and image inpainting is presented in this work . | {
"relations": {
"used for": [
{
"head": {
"text": "exemplar-based framework",
"start": 6,
"end": 30
},
"tail": {
"text": "image completion",
"start": 40,
"end": 56
}
},
{
"head": {
"tex... | [] |
Contrary to existing greedy techniques , these tasks are posed in the form of a discrete global optimization problem with a well defined objective function . | {
"relations": {
"compare": [
{
"head": {
"text": "greedy techniques",
"start": 21,
"end": 38
},
"tail": {
"text": "tasks",
"start": 47,
"end": 52
}
}
],
"feature of": [
{
"head": {
... | [] |
For solving this problem a novel optimization scheme , called Priority-BP , is proposed which carries two very important extensions over standard belief propagation -LRB- BP -RRB- : '' priority-based message scheduling '' and '' dynamic label pruning '' . | {
"relations": {
"used for": [
{
"head": {
"text": "optimization scheme",
"start": 33,
"end": 52
},
"tail": {
"text": "problem",
"start": 17,
"end": 24
}
},
{
"head": {
"text": "belief p... | [] |
These two extensions work in cooperation to deal with the intolerable computational cost of BP caused by the huge number of existing labels . | {
"relations": {
"used for": [
{
"head": {
"text": "extensions",
"start": 10,
"end": 20
},
"tail": {
"text": "intolerable computational cost of BP",
"start": 58,
"end": 94
}
}
]
}
} | [] |
Moreover , both extensions are generic and can therefore be applied to any MRF energy function as well . | {
"relations": {
"used for": [
{
"head": {
"text": "extensions",
"start": 16,
"end": 26
},
"tail": {
"text": "MRF energy function",
"start": 75,
"end": 94
}
}
]
}
} | [] |
The effectiveness of our method is demonstrated on a wide variety of image completion examples . | {
"relations": {
"used for": [
{
"head": {
"text": "image completion examples",
"start": 69,
"end": 94
},
"tail": {
"text": "method",
"start": 25,
"end": 31
}
}
]
}
} | [] |
In this paper , we compare the relative effects of segment order , segmentation and segment contiguity on the retrieval performance of a translation memory system . | {
"relations": {
"conjunction": [
{
"head": {
"text": "segment order",
"start": 51,
"end": 64
},
"tail": {
"text": "segmentation",
"start": 67,
"end": 79
}
},
{
"head": {
"text": "segmen... | [] |
We take a selection of both bag-of-words and segment order-sensitive string comparison methods , and run each over both character - and word-segmented data , in combination with a range of local segment contiguity models -LRB- in the form of N-grams -RRB- . | {
"relations": {
"used for": [
{
"head": {
"text": "character - and word-segmented data",
"start": 120,
"end": 155
},
"tail": {
"text": "bag-of-words and segment order-sensitive string comparison methods",
"start": 28,
"end"... | [] |
Over two distinct datasets , we find that indexing according to simple character bigrams produces a retrieval accuracy superior to any of the tested word N-gram models . | {
"relations": {
"used for": [
{
"head": {
"text": "character bigrams",
"start": 71,
"end": 88
},
"tail": {
"text": "indexing",
"start": 42,
"end": 50
}
}
],
"compare": [
{
"head": {
... | [] |
Further , in their optimum configuration , bag-of-words methods are shown to be equivalent to segment order-sensitive methods in terms of retrieval accuracy , but much faster . | {
"relations": {
"compare": [
{
"head": {
"text": "bag-of-words methods",
"start": 43,
"end": 63
},
"tail": {
"text": "segment order-sensitive methods",
"start": 94,
"end": 125
}
}
],
"evaluate for": [
... | [] |
In this paper we show how two standard outputs from information extraction -LRB- IE -RRB- systems - named entity annotations and scenario templates - can be used to enhance access to text collections via a standard text browser . | {
"relations": {
"used for": [
{
"head": {
"text": "outputs",
"start": 39,
"end": 46
},
"tail": {
"text": "text collections",
"start": 183,
"end": 199
}
},
{
"head": {
"text": "text brow... | [] |
We describe how this information is used in a prototype system designed to support information workers ' access to a pharmaceutical news archive as part of their industry watch function . | {
"relations": {
"used for": [
{
"head": {
"text": "prototype system",
"start": 46,
"end": 62
},
"tail": {
"text": "pharmaceutical news archive",
"start": 117,
"end": 144
}
}
]
}
} | [] |
We also report results of a preliminary , qualitative user evaluation of the system , which while broadly positive indicates further work needs to be done on the interface to make users aware of the increased potential of IE-enhanced text browsers . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "qualitative user evaluation",
"start": 42,
"end": 69
},
"tail": {
"text": "system",
"start": 77,
"end": 83
}
}
]
}
} | [] |
We present a new model-based bundle adjustment algorithm to recover the 3D model of a scene/object from a sequence of images with unknown motions . | {
"relations": {
"used for": [
{
"head": {
"text": "model-based bundle adjustment algorithm",
"start": 17,
"end": 56
},
"tail": {
"text": "3D model",
"start": 72,
"end": 80
}
},
{
"head": {
... | [] |
Instead of representing scene/object by a collection of isolated 3D features -LRB- usually points -RRB- , our algorithm uses a surface controlled by a small set of parameters . | {
"relations": {
"used for": [
{
"head": {
"text": "surface",
"start": 127,
"end": 134
},
"tail": {
"text": "algorithm",
"start": 110,
"end": 119
}
}
]
}
} | [] |
Compared with previous model-based approaches , our approach has the following advantages . | {
"relations": {
"compare": [
{
"head": {
"text": "model-based approaches",
"start": 23,
"end": 45
},
"tail": {
"text": "approach",
"start": 35,
"end": 43
}
}
]
}
} | [] |
First , instead of using the model space as a regular-izer , we directly use it as our search space , thus resulting in a more elegant formulation with fewer unknowns and fewer equations . | {
"relations": {
"used for": [
{
"head": {
"text": "model space",
"start": 29,
"end": 40
},
"tail": {
"text": "regular-izer",
"start": 46,
"end": 58
}
},
{
"head": {
"text": "it",
... | [] |
Third , regarding face modeling , we use a very small set of face metrics -LRB- meaningful deformations -RRB- to parame-terize the face geometry , resulting in a smaller search space and a better posed system . | {
"relations": {
"used for": [
{
"head": {
"text": "face metrics",
"start": 61,
"end": 73
},
"tail": {
"text": "face modeling",
"start": 18,
"end": 31
}
},
{
"head": {
"text": "face metr... | [] |
Experiments with both synthetic and real data show that this new algorithm is faster , more accurate and more stable than existing ones . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "synthetic and real data",
"start": 22,
"end": 45
},
"tail": {
"text": "algorithm",
"start": 65,
"end": 74
}
},
{
"head": {
"text":... | [] |
This paper presents an approach to the unsupervised learning of parts of speech which uses both morphological and syntactic information . | {
"relations": {
"used for": [
{
"head": {
"text": "approach",
"start": 23,
"end": 31
},
"tail": {
"text": "unsupervised learning of parts of speech",
"start": 39,
"end": 79
}
},
{
"head": {
... | [] |
While the model is more complex than those which have been employed for unsupervised learning of POS tags in English , which use only syntactic information , the variety of languages in the world requires that we consider morphology as well . | {
"relations": {
"compare": [
{
"head": {
"text": "model",
"start": 10,
"end": 15
},
"tail": {
"text": "those",
"start": 37,
"end": 42
}
}
],
"used for": [
{
"head": {
"text": "t... | [] |
In many languages , morphology provides better clues to a word 's category than word order . | {
"relations": {
"compare": [
{
"head": {
"text": "morphology",
"start": 20,
"end": 30
},
"tail": {
"text": "word order",
"start": 80,
"end": 90
}
}
]
}
} | [] |
We present the computational model for POS learning , and present results for applying it to Bulgarian , a Slavic language with relatively free word order and rich morphology . | {
"relations": {
"used for": [
{
"head": {
"text": "computational model",
"start": 15,
"end": 34
},
"tail": {
"text": "POS learning",
"start": 39,
"end": 51
}
},
{
"head": {
"text": "Bul... | [] |
In MT , the widely used approach is to apply a Chinese word segmenter trained from manually annotated data , using a fixed lexicon . | {
"relations": {
"used for": [
{
"head": {
"text": "Chinese word segmenter",
"start": 47,
"end": 69
},
"tail": {
"text": "MT",
"start": 3,
"end": 5
}
},
{
"head": {
"text": "manually ann... | [] |
Such word segmentation is not necessarily optimal for translation . | {
"relations": {
"used for": [
{
"head": {
"text": "word segmentation",
"start": 5,
"end": 22
},
"tail": {
"text": "translation",
"start": 54,
"end": 65
}
}
]
}
} | [] |
We propose a Bayesian semi-supervised Chinese word segmentation model which uses both monolingual and bilingual information to derive a segmentation suitable for MT . | {
"relations": {
"used for": [
{
"head": {
"text": "Bayesian semi-supervised Chinese word segmentation model",
"start": 13,
"end": 69
},
"tail": {
"text": "segmentation",
"start": 51,
"end": 63
}
},
{
... | [] |
Experiments show that our method improves a state-of-the-art MT system in a small and a large data environment . | {
"relations": {
"compare": [
{
"head": {
"text": "method",
"start": 26,
"end": 32
},
"tail": {
"text": "MT system",
"start": 61,
"end": 70
}
}
]
}
} | [] |
In this paper we compare two competing approaches to part-of-speech tagging , statistical and constraint-based disambiguation , using French as our test language . | {
"relations": {
"used for": [
{
"head": {
"text": "approaches",
"start": 39,
"end": 49
},
"tail": {
"text": "part-of-speech tagging",
"start": 53,
"end": 75
}
},
{
"head": {
"text": "Fr... | [] |
We imposed a time limit on our experiment : the amount of time spent on the design of our constraint system was about the same as the time we used to train and test the easy-to-implement statistical model . | {
"relations": {
"compare": [
{
"head": {
"text": "constraint system",
"start": 90,
"end": 107
},
"tail": {
"text": "statistical model",
"start": 187,
"end": 204
}
}
]
}
} | [] |
The accuracy of the statistical method is reasonably good , comparable to taggers for English . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "accuracy",
"start": 4,
"end": 12
},
"tail": {
"text": "statistical method",
"start": 20,
"end": 38
}
},
{
"head": {
"text": "accur... | [] |
Structured-light methods actively generate geometric correspondence data between projectors and cameras in order to facilitate robust 3D reconstruction . | {
"relations": {
"used for": [
{
"head": {
"text": "Structured-light methods",
"start": 0,
"end": 24
},
"tail": {
"text": "geometric correspondence data",
"start": 43,
"end": 72
}
},
{
"head": {
... | [] |
In this paper , we present Photogeometric Structured Light whereby a standard structured light method is extended to include photometric methods . | {
"relations": {
"part of": [
{
"head": {
"text": "structured light method",
"start": 78,
"end": 101
},
"tail": {
"text": "Photogeometric Structured Light",
"start": 27,
"end": 58
}
},
{
"head": {... | [] |
Photometric processing serves the double purpose of increasing the amount of recovered surface detail and of enabling the structured-light setup to be robustly self-calibrated . | {
"relations": {
"used for": [
{
"head": {
"text": "Photometric processing",
"start": 0,
"end": 22
},
"tail": {
"text": "recovered surface detail",
"start": 77,
"end": 101
}
},
{
"head": {
... | [] |
Further , our framework uses a photogeometric optimization that supports the simultaneous use of multiple cameras and projectors and yields a single and accurate multi-view 3D model which best complies with photometric and geometric data . | {
"relations": {
"used for": [
{
"head": {
"text": "photogeometric optimization",
"start": 31,
"end": 58
},
"tail": {
"text": "framework",
"start": 14,
"end": 23
}
},
{
"head": {
"text":... | [] |
In this paper , a discrimination and robustness oriented adaptive learning procedure is proposed to deal with the task of syntactic ambiguity resolution . | {
"relations": {
"used for": [
{
"head": {
"text": "adaptive learning procedure",
"start": 57,
"end": 84
},
"tail": {
"text": "syntactic ambiguity resolution",
"start": 122,
"end": 152
}
}
]
}
} | [] |
Owing to the problem of insufficient training data and approximation error introduced by the language model , traditional statistical approaches , which resolve ambiguities by indirectly and implicitly using maximum likelihood method , fail to achieve high performance in real applications . | {
"relations": {
"conjunction": [
{
"head": {
"text": "insufficient training data",
"start": 24,
"end": 50
},
"tail": {
"text": "approximation error",
"start": 55,
"end": 74
}
}
],
"used for": [
{... | [] |
The accuracy rate of syntactic disambiguation is raised from 46.0 % to 60.62 % by using this novel approach . | {
"relations": {
"evaluate for": [
{
"head": {
"text": "accuracy rate",
"start": 4,
"end": 17
},
"tail": {
"text": "syntactic disambiguation",
"start": 21,
"end": 45
}
},
{
"head": {
"te... | [] |
This paper presents a new approach to statistical sentence generation in which alternative phrases are represented as packed sets of trees , or forests , and then ranked statistically to choose the best one . | {
"relations": {
"used for": [
{
"head": {
"text": "approach",
"start": 26,
"end": 34
},
"tail": {
"text": "statistical sentence generation",
"start": 38,
"end": 69
}
}
]
}
} | [] |
It also facilitates more efficient statistical ranking than a previous approach to statistical generation . | {
"relations": {
"used for": [
{
"head": {
"text": "It",
"start": 0,
"end": 2
},
"tail": {
"text": "statistical ranking",
"start": 35,
"end": 54
}
},
{
"head": {
"text": "approach",
... | [] |
An efficient ranking algorithm is described , together with experimental results showing significant improvements over simple enumeration or a lattice-based approach . | {
"relations": {
"compare": [
{
"head": {
"text": "ranking algorithm",
"start": 13,
"end": 30
},
"tail": {
"text": "enumeration",
"start": 126,
"end": 137
}
},
{
"head": {
"text": "ranki... | [] |
This article deals with the interpretation of conceptual operations underlying the communicative use of natural language -LRB- NL -RRB- within the Structured Inheritance Network -LRB- SI-Nets -RRB- paradigm . | {
"relations": {
"used for": [
{
"head": {
"text": "natural language -LRB- NL -RRB-",
"start": 104,
"end": 135
},
"tail": {
"text": "Structured Inheritance Network -LRB- SI-Nets -RRB- paradigm",
"start": 147,
"end": 206
... | [] |
The operations are reduced to functions of a formal language , thus changing the level of abstraction of the operations to be performed on SI-Nets . | {
"relations": {
"used for": [
{
"head": {
"text": "operations",
"start": 4,
"end": 14
},
"tail": {
"text": "SI-Nets",
"start": 139,
"end": 146
}
}
]
}
} | [] |
In this sense , operations on SI-Nets are not merely isomorphic to single epistemological objects , but can be viewed as a simulation of processes on a different level , that pertaining to the conceptual system of NL . | {
"relations": {
"used for": [
{
"head": {
"text": "operations",
"start": 16,
"end": 26
},
"tail": {
"text": "SI-Nets",
"start": 30,
"end": 37
}
},
{
"head": {
"text": "NL",
"s... | [] |
For this purpose , we have designed a version of KL-ONE which represents the epistemological level , while the new experimental language , KL-Conc , represents the conceptual level . | {
"relations": {
"compare": [
{
"head": {
"text": "KL-ONE",
"start": 49,
"end": 55
},
"tail": {
"text": "KL-Conc",
"start": 139,
"end": 146
}
}
],
"feature of": [
{
"head": {
"te... | [] |
We present an algorithm for calibrated camera relative pose estimation from lines . | {
"relations": {
"used for": [
{
"head": {
"text": "algorithm",
"start": 14,
"end": 23
},
"tail": {
"text": "calibrated camera relative pose estimation",
"start": 28,
"end": 70
}
}
]
}
} | [] |
We evaluate the performance of the algorithm using synthetic and real data . | {
"relations": {
"used for": [
{
"head": {
"text": "synthetic and real data",
"start": 51,
"end": 74
},
"tail": {
"text": "algorithm",
"start": 35,
"end": 44
}
}
]
}
} | [] |
The intended use of the algorithm is with robust hypothesize-and-test frameworks such as RANSAC . | {
"relations": {
"conjunction": [
{
"head": {
"text": "algorithm",
"start": 24,
"end": 33
},
"tail": {
"text": "hypothesize-and-test frameworks",
"start": 49,
"end": 80
}
}
],
"hyponym of": [
{
... | [] |
Our approach is suitable for urban and indoor environments where most lines are either parallel or orthogonal to each other . | {
"relations": {
"used for": [
{
"head": {
"text": "approach",
"start": 4,
"end": 12
},
"tail": {
"text": "urban and indoor environments",
"start": 29,
"end": 58
}
}
]
}
} | [] |
In this paper , we present a fully automated extraction system , named IntEx , to identify gene and protein interactions in biomedical text . | {
"relations": {
"used for": [
{
"head": {
"text": "fully automated extraction system",
"start": 29,
"end": 62
},
"tail": {
"text": "gene and protein interactions",
"start": 91,
"end": 120
}
},
{
... | [] |
Then , tagging biological entities with the help of biomedical and linguistic ontologies . | {
"relations": {
"used for": [
{
"head": {
"text": "biomedical and linguistic ontologies",
"start": 52,
"end": 88
},
"tail": {
"text": "biological entities",
"start": 15,
"end": 34
}
}
]
}
} | [] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.