1 |
MCSQ Translation Models (en-ru) (v1.0)
|
|
Variš, Dušan. - : Charles University, Faculty of Mathematics and Physics, Institute of Formal and Applied Linguistics (UFAL), 2022
|
|
BASE
|
|
Show details
|
|
2 |
MCSQ Translation Models (en-de) (v1.0)
|
|
Variš, Dušan. - : Charles University, Faculty of Mathematics and Physics, Institute of Formal and Applied Linguistics (UFAL), 2022
|
|
BASE
|
|
Show details
|
|
3 |
Evaluation computergestützter Verfahren der Emotionsklassifikation für deutschsprachige Dramen um 1800 ...
|
|
|
|
BASE
|
|
Show details
|
|
4 |
Evaluation computergestützter Verfahren der Emotionsklassifikation für deutschsprachige Dramen um 1800 ...
|
|
|
|
BASE
|
|
Show details
|
|
5 |
Transformer-Based Abstractive Summarization for Reddit and Twitter: Single Posts vs. Comment Pools in Three Languages
|
|
|
|
In: Future Internet; Volume 14; Issue 3; Pages: 69 (2022)
|
|
BASE
|
|
Show details
|
|
6 |
Speech Enhancement by Multiple Propagation through the Same Neural Network
|
|
|
|
In: Sensors; Volume 22; Issue 7; Pages: 2440 (2022)
|
|
BASE
|
|
Show details
|
|
7 |
FedQAS: Privacy-Aware Machine Reading Comprehension with Federated Learning
|
|
|
|
In: Applied Sciences; Volume 12; Issue 6; Pages: 3130 (2022)
|
|
Abstract:
Machine reading comprehension (MRC) of text data is a challenging task in Natural Language Processing (NLP), with a lot of ongoing research fueled by the release of the Stanford Question Answering Dataset (SQuAD) and Conversational Question Answering (CoQA). It is considered to be an effort to teach computers how to “understand” a text, and then to be able to answer questions about it using deep learning. However, until now, large-scale training on private text data and knowledge sharing has been missing for this NLP task. Hence, we present FedQAS, a privacy-preserving machine reading system capable of leveraging large-scale private data without the need to pool those datasets in a central location. The proposed approach combines transformer models and federated learning technologies. The system is developed using the FEDn framework and deployed as a proof-of-concept alliance initiative. FedQAS is flexible, language-agnostic, and allows intuitive participation and execution of local model training. In addition, we present the architecture and implementation of the system, as well as provide a reference evaluation based on the SQuAD dataset, to showcase how it overcomes data privacy issues and enables knowledge sharing between alliance members in a Federated learning setting.
|
|
Keyword:
data privacy; federated learning; machine reading comprehension; natural language processing; question answering; transformer
|
|
URL: https://doi.org/10.3390/app12063130
|
|
BASE
|
|
Hide details
|
|
8 |
A Pipeline Approach to Context-Aware Handwritten Text Recognition
|
|
|
|
In: Applied Sciences; Volume 12; Issue 4; Pages: 1870 (2022)
|
|
BASE
|
|
Show details
|
|
9 |
Correcting Diacritics and Typos with a ByT5 Transformer Model
|
|
|
|
In: Applied Sciences; Volume 12; Issue 5; Pages: 2636 (2022)
|
|
BASE
|
|
Show details
|
|
10 |
Research on Named Entity Recognition Methods in Chinese Forest Disease Texts
|
|
|
|
In: Applied Sciences; Volume 12; Issue 8; Pages: 3885 (2022)
|
|
BASE
|
|
Show details
|
|
11 |
Cross-Lingual Transfer Learning for Arabic Task-Oriented Dialogue Systems Using Multilingual Transformer Model mT5
|
|
|
|
In: Mathematics; Volume 10; Issue 5; Pages: 746 (2022)
|
|
BASE
|
|
Show details
|
|
12 |
AraConv: Developing an Arabic Task-Oriented Dialogue System Using Multi-Lingual Transformer Model mT5
|
|
|
|
In: Applied Sciences; Volume 12; Issue 4; Pages: 1881 (2022)
|
|
BASE
|
|
Show details
|
|
13 |
Retrieval-Based Transformer Pseudocode Generation
|
|
|
|
In: Mathematics; Volume 10; Issue 4; Pages: 604 (2022)
|
|
BASE
|
|
Show details
|
|
14 |
Hebrew Transformed: Machine Translation of Hebrew Using the Transformer Architecture
|
|
|
|
BASE
|
|
Show details
|
|
15 |
English machine reading comprehension: new approaches to answering multiple-choice questions
|
|
Dzendzik, Daria. - : Dublin City University. School of Computing, 2021. : Dublin City University. ADAPT, 2021
|
|
In: Dzendzik, Daria (2021) English machine reading comprehension: new approaches to answering multiple-choice questions. PhD thesis, Dublin City University. (2021)
|
|
BASE
|
|
Show details
|
|
16 |
Transformer versus LSTM Language Models Trained on Uncertain ASR Hypotheses in Limited Data Scenarios
|
|
|
|
In: https://hal.inria.fr/hal-03362828 ; 2021 (2021)
|
|
BASE
|
|
Show details
|
|
17 |
Simulating reading mistakes for child speech Transformer-based phone recognition
|
|
|
|
In: Annual Conference of the International Speech Communication Association (INTERSPEECH) ; https://hal.archives-ouvertes.fr/hal-03257870 ; Annual Conference of the International Speech Communication Association (INTERSPEECH), Aug 2021, Brno, Czech Republic (2021)
|
|
BASE
|
|
Show details
|
|
18 |
Breaking Down the Invisible Wall of Informal Fallacies in Online Discussions
|
|
|
|
In: ACL-IJCNLP 2021 - Joint Conference of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing ; https://hal.inria.fr/hal-03351649 ; ACL-IJCNLP 2021 - Joint Conference of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, Aug 2021, Online, France ; https://2021.aclweb.org/ (2021)
|
|
BASE
|
|
Show details
|
|
19 |
End-to-end acoustic modelling for phone recognition of young readers
|
|
|
|
In: ISSN: 0167-6393 ; EISSN: 1872-7182 ; Speech Communication ; https://hal.archives-ouvertes.fr/hal-03373156 ; Speech Communication, Elsevier : North-Holland, 2021, 134, pp.71-84. ⟨10.1016/j.specom.2021.08.003⟩ ; https://www.sciencedirect.com/science/article/pii/S0167639321000959?via%3Dihub (2021)
|
|
BASE
|
|
Show details
|
|
20 |
Multitask Transformer Model-based Fintech Customer Service Chatbot NLU System with DECO-LGG SSP-based Data ; DECO-LGG 반자동 증강 학습데이터 활용 멀티태스크 트랜스포머 모델 기반 핀테크 CS 챗봇 NLU 시스템
|
|
|
|
In: Annual Conference on Human and Language Technology ; https://hal.archives-ouvertes.fr/hal-03603903 ; Annual Conference on Human and Language Technology, Oct 2021, Séoul, South Korea. pp.461-466 ; http://www.koreascience.or.kr/journal/OOGHAK.page (2021)
|
|
BASE
|
|
Show details
|
|
|
|