The Experts below are selected from a list of 360 Experts worldwide ranked by ideXlab platform
Iryna Gurevych - One of the best experts on this subject based on the ideXlab platform.
-
Text Processing like humans do visually attacking and shielding nlp systems
arXiv: Computation and Language, 2019Co-Authors: Steffen Eger, Gozde Gul şahin, Andreas Ruckle, Jiung Lee, Claudia Schulz, Mohsen Mesgar, Krishnkant Swarnkar, Edwin Simpson, Iryna GurevychAbstract:Visual modifications to Text are often used to obfuscate offensive comments in social media (e.g., "!d10t") or as a writing style ("1337" in "leet speak"), among other scenarios. We consider this as a new type of adversarial attack in NLP, a setting to which humans are very robust, as our experiments with both simple and more difficult visual input perturbations demonstrate. We then investigate the impact of visual adversarial attacks on current NLP systems on character-, word-, and sentence-level tasks, showing that both neural and non-neural models are, in contrast to humans, extremely sensitive to such attacks, suffering performance decreases of up to 82\%. We then explore three shielding methods---visual character embeddings, adversarial training, and rule-based recovery---which substantially improve the robustness of the models. However, the shielding methods still fall behind performances achieved in non-attack scenarios, which demonstrates the difficulty of dealing with visual attacks.
-
Text Processing like humans do visually attacking and shielding nlp systems
North American Chapter of the Association for Computational Linguistics, 2019Co-Authors: Steffen Eger, Gozde Gul şahin, Andreas Ruckle, Jiung Lee, Claudia Schulz, Mohsen Mesgar, Krishnkant Swarnkar, Edwin Simpson, Iryna GurevychAbstract:Visual modifications to Text are often used to obfuscate offensive comments in social media (e.g., “!d10t”) or as a writing style (“1337” in “leet speak”), among other scenarios. We consider this as a new type of adversarial attack in NLP, a setting to which humans are very robust, as our experiments with both simple and more difficult visual perturbations demonstrate. We investigate the impact of visual adversarial attacks on current NLP systems on character-, word-, and sentence-level tasks, showing that both neural and non-neural models are, in contrast to humans, extremely sensitive to such attacks, suffering performance decreases of up to 82%. We then explore three shielding methods—visual character embeddings, adversarial training, and rule-based recovery—which substantially improve the robustness of the models. However, the shielding methods still fall behind performances achieved in non-attack scenarios, which demonstrates the difficulty of dealing with visual attacks.
John Richardson - One of the best experts on this subject based on the ideXlab platform.
-
sentencepiece a simple and language independent subword tokenizer and detokenizer for neural Text Processing
arXiv: Computation and Language, 2018Co-Authors: Taku Kudo, John RichardsonAbstract:This paper describes SentencePiece, a language-independent subword tokenizer and detokenizer designed for Neural-based Text Processing, including Neural Machine Translation. It provides open-source C++ and Python implementations for subword units. While existing subword segmentation tools assume that the input is pre-tokenized into word sequences, SentencePiece can train subword models directly from raw sentences, which allows us to make a purely end-to-end and language independent system. We perform a validation experiment of NMT on English-Japanese machine translation, and find that it is possible to achieve comparable accuracy to direct subword training from raw sentences. We also compare the performance of subword training and segmentation with various configurations. SentencePiece is available under the Apache 2 license at this https URL.
-
sentencepiece a simple and language independent subword tokenizer and detokenizer for neural Text Processing
Empirical Methods in Natural Language Processing, 2018Co-Authors: Taku Kudo, John RichardsonAbstract:This paper describes SentencePiece, a language-independent subword tokenizer and detokenizer designed for Neural-based Text Processing, including Neural Machine Translation. It provides open-source C++ and Python implementations for subword units. While existing subword segmentation tools assume that the input is pre-tokenized into word sequences, SentencePiece can train subword models directly from raw sentences, which allows us to make a purely end-to-end and language independent system. We perform a validation experiment of NMT on English-Japanese machine translation, and find that it is possible to achieve comparable accuracy to direct subword training from raw sentences. We also compare the performance of subword training and segmentation with various configurations. SentencePiece is available under the Apache 2 license at https://github.com/google/sentencepiece.
Tetsuya Nasukawa - One of the best experts on this subject based on the ideXlab platform.
-
full Text Processing improving a practical nlp system based on surface information within the conText
International Conference on Computational Linguistics, 1996Co-Authors: Tetsuya NasukawaAbstract:Rich information for resolving ambiguities in sentence analysis, including various conText-dependent problems, can be obtained by analyzing a simple set of parsed trees of each sentence in a Text without constructing a precise model of the conText through deep semantic analysis. Thus, Processing a group of sentences together makes it possible to improve the accuracy of a practical natural language Processing (NLP) system such as a machine translation system. In this paper, we describe a simple conText model consisting of parsed trees of each sentence in a Text, and its effectiveness for handling various problems in NLP such as the resolution of structural ambiguities, pronoun referents, and the focus of focusing subjects (e.g. also and only), as well as for adding supplementary phrases to some elliptical sentences.
Steffen Eger - One of the best experts on this subject based on the ideXlab platform.
-
Text Processing like humans do visually attacking and shielding nlp systems
arXiv: Computation and Language, 2019Co-Authors: Steffen Eger, Gozde Gul şahin, Andreas Ruckle, Jiung Lee, Claudia Schulz, Mohsen Mesgar, Krishnkant Swarnkar, Edwin Simpson, Iryna GurevychAbstract:Visual modifications to Text are often used to obfuscate offensive comments in social media (e.g., "!d10t") or as a writing style ("1337" in "leet speak"), among other scenarios. We consider this as a new type of adversarial attack in NLP, a setting to which humans are very robust, as our experiments with both simple and more difficult visual input perturbations demonstrate. We then investigate the impact of visual adversarial attacks on current NLP systems on character-, word-, and sentence-level tasks, showing that both neural and non-neural models are, in contrast to humans, extremely sensitive to such attacks, suffering performance decreases of up to 82\%. We then explore three shielding methods---visual character embeddings, adversarial training, and rule-based recovery---which substantially improve the robustness of the models. However, the shielding methods still fall behind performances achieved in non-attack scenarios, which demonstrates the difficulty of dealing with visual attacks.
-
Text Processing like humans do visually attacking and shielding nlp systems
North American Chapter of the Association for Computational Linguistics, 2019Co-Authors: Steffen Eger, Gozde Gul şahin, Andreas Ruckle, Jiung Lee, Claudia Schulz, Mohsen Mesgar, Krishnkant Swarnkar, Edwin Simpson, Iryna GurevychAbstract:Visual modifications to Text are often used to obfuscate offensive comments in social media (e.g., “!d10t”) or as a writing style (“1337” in “leet speak”), among other scenarios. We consider this as a new type of adversarial attack in NLP, a setting to which humans are very robust, as our experiments with both simple and more difficult visual perturbations demonstrate. We investigate the impact of visual adversarial attacks on current NLP systems on character-, word-, and sentence-level tasks, showing that both neural and non-neural models are, in contrast to humans, extremely sensitive to such attacks, suffering performance decreases of up to 82%. We then explore three shielding methods—visual character embeddings, adversarial training, and rule-based recovery—which substantially improve the robustness of the models. However, the shielding methods still fall behind performances achieved in non-attack scenarios, which demonstrates the difficulty of dealing with visual attacks.
Gobinda Chowdhury - One of the best experts on this subject based on the ideXlab platform.
-
Natural Language Processing
Language, 2003Co-Authors: Gobinda G Chowdhury, Gobinda ChowdhuryAbstract:Beginning with the basic issues of NLP, this chapter aims to chart the major research activities in this area since the last ARIST Chapter in 1996 (Haas, 1996), including: (i) natural language Text Processing systems - Text summarization, information extraction, information retrieval, etc., including domain-specific applications; (ii) natural language interfaces; (iii) NLP in the conText of www and digital libraries ; and (iv) evaluation of NLP systems.
