Page: 1 2 3 4 5 6 7 8 9... 83
81 |
Idiomatic Expression Identification using Semantic Compatibility
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 1546-1562 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
82 |
KEPLER: A Unified Model for Knowledge Embedding and Pre-trained Language Representation
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 176-194 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
83 |
Reducing Confusion in Active Learning for Part-Of-Speech Tagging
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 1-16 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
84 |
Differentiable Subset Pruning of Transformer Heads
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 1442-1459 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
85 |
Compressing Large-Scale Transformer-Based Models: A Case Study on BERT
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 1061-1080 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
86 |
Parameter Space Factorization for Zero-Shot Learning across Tasks and Languages
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 410-428 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
87 |
Data-to-text Generation with Macro Planning
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 510-527 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
88 |
Structured Self-Supervised Pretraining for Commonsense Knowledge Graph Completion
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 1268-1284 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
89 |
RYANSQL: Recursively Applying Sketch-based Slot Fillings for Complex Text-to-SQL in Cross-Domain Databases
|
|
|
|
In: Computational Linguistics, Vol 47, Iss 2, Pp 309-332 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
90 |
Narrative Question Answering with Cutting-Edge Open-Domain QA Techniques: A Comprehensive Study
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 1032-1046 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
91 |
Maintaining Common Ground in Dynamic Environments
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 995-1011 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
92 |
Infusing Finetuning with Semantic Dependencies
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 226-242 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
93 |
On Generative Spoken Language Modeling from Raw Audio
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 1336-1354 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
94 |
Pretraining the Noisy Channel Model for Task-Oriented Dialogue
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 657-674 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
95 |
Approximating Probabilistic Models as Weighted Finite Automata
|
|
|
|
In: Computational Linguistics, Vol 47, Iss 2, Pp 221-254 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
96 |
Sensitivity as a Complexity Measure for Sequence Classification Tasks
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 891-908 (2021) (2021)
|
|
Abstract:
AbstractWe introduce a theoretical framework for understanding and predicting the complexity of sequence classification tasks, using a novel extension of the theory of Boolean function sensitivity. The sensitivity of a function, given a distribution over input sequences, quantifies the number of disjoint subsets of the input sequence that can each be individually changed to change the output. We argue that standard sequence classification methods are biased towards learning low-sensitivity functions, so that tasks requiring high sensitivity are more difficult. To that end, we show analytically that simple lexical classifiers can only express functions of bounded sensitivity, and we show empirically that low-sensitivity functions are easier to learn for LSTMs. We then estimate sensitivity on 15 NLP tasks, finding that sensitivity is higher on challenging tasks collected in GLUE than on simple text classification tasks, and that sensitivity predicts the performance both of simple lexical classifiers and of vanilla BiLSTMs without pretrained contextualized embeddings. Within a task, sensitivity predicts which inputs are hard for such simple models. Our results suggest that the success of massively pretrained contextual representations stems in part because they provide representations from which information can be extracted by low-sensitivity decoders.
|
|
Keyword:
Computational linguistics. Natural language processing; P98-98.5
|
|
URL: https://doaj.org/article/958ef3445dbd4c7dbaea2d7c380df722 https://doi.org/10.1162/tacl_a_00403
|
|
BASE
|
|
Hide details
|
|
97 |
Unsupervised Learning of KB Queries in Task-Oriented Dialogs
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 374-390 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
98 |
<scp>ParsiNLU</scp>: A Suite of Language Understanding Challenges for Persian
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 1147-1162 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
99 |
Adaptive Semiparametric Language Models
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 362-373 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
100 |
Strong Equivalence of TAG and CCG
|
|
|
|
In: Transactions of the Association for Computational Linguistics, Vol 9, Pp 707-720 (2021) (2021)
|
|
BASE
|
|
Show details
|
|
Page: 1 2 3 4 5 6 7 8 9... 83
|
|