Zobrazeno 1 - 10
of 68
pro vyhledávání: '"Kentaro Torisawa"'
Publikováno v:
Journal of Natural Language Processing. 28:1299-1306
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::1ae19bb1d6a3da6985c815bab56a555c
https://doi.org/10.5715/jnlp.28.1299
https://doi.org/10.5715/jnlp.28.1299
Publikováno v:
IPDPS
This work proposes RaNNC (Rapid Neural Network Connector) as middleware for automatic hybrid parallelism. In recent deep learning research, as exemplified by T5 and GPT-3, the size of neural network models continues to grow. Since such models do not
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::7c183e68fdd9425bdb9febbbdc4979ed
Publikováno v:
ACL (1)
In this paper, we propose a method for why-question answering (why-QA) that uses an adversarial learning framework. Existing why-QA methods retrieve “answer passages” that usually consist of several sentences. These multi-sentence passages contai
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::3b2d7e86a7394cb86eb5a212951922eb
https://doi.org/10.18653/v1/p19-1414
https://doi.org/10.18653/v1/p19-1414
Publikováno v:
EMNLP/IJCNLP (1)
We propose new BERT-based methods for recognizing event causality such as “smoke cigarettes” –> “die of lung cancer” written in web texts. In our methods, we grasp each annotator’s policy by training multiple classifiers, each of which pr
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::ee4f9f88bb12f61f35851d40849a6391
https://doi.org/10.18653/v1/d19-1590
https://doi.org/10.18653/v1/d19-1590
Publikováno v:
ICPADS
Research and development of deep learning (DL) applications often involves exhaustive trial-and-error, which demands that shared computational resources, especially GPUs, be efficiently allocated. Most DL tasks are moldable or malleable (i.e., the nu
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::2cabd6e5939697c802ee12787574c9da
https://doi.org/10.1109/padsw.2018.8644536
https://doi.org/10.1109/padsw.2018.8644536
Publikováno v:
Proceedings of the AAAI Conference on Artificial Intelligence. 32
This paper proposes a neural network-based method for generating compact answers to open-domain why-questions (e.g., "Why was Mr. Trump elected as the president of the US?"). Unlike factoid question answering methods that provide short text spans as
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::9d4398f8319b76a2722fa4886fb9af32
https://doi.org/10.1609/aaai.v32i1.12064
https://doi.org/10.1609/aaai.v32i1.12064
Publikováno v:
IPDPS
Large-scale data analysis applications are becoming more and more prevalent in a wide variety of areas. These applications are composed of many currently available programs called analysis components. Thousands of analysis component processes are orc
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::dfcb1fe433e71d145e11a0c8e7a1d3ac
https://doi.org/10.1109/ipdps.2017.89
https://doi.org/10.1109/ipdps.2017.89
Autor:
Canasai Kruengkrai, Kentaro Torisawa, Chikara Hashimoto, Julien Kloetzer, Jong-Hoon Oh, Masahiro Tanaka
Publikováno v:
Proceedings of the AAAI Conference on Artificial Intelligence. 31
We propose a method for recognizing such event causalities as "smoke cigarettes" → "die of lung cancer" using background knowledge taken from web texts as well as original sentences from which candidates for the causalities were extracted. We retri
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::0775e6a997f8d90db1a0ab9e4024b906
https://doi.org/10.1609/aaai.v31i1.11005
https://doi.org/10.1609/aaai.v31i1.11005
Publikováno v:
IEICE Communications Society Magazine. 8:18-25
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::504b569deffd58739a056559c80f2729
https://doi.org/10.1587/bplus.8.18
https://doi.org/10.1587/bplus.8.18
Publikováno v:
Medinfo; 2019, p423-427, 5p