2 Search Results for "Dyer, Chris"


Document
From Characters to Understanding Natural Language (C2NLU): Robust End-to-End Deep Learning for NLP (Dagstuhl Seminar 17042)

Authors: Phil Blunsom, Kyunghyun Cho, Chris Dyer, and Hinrich Schütze

Published in: Dagstuhl Reports, Volume 7, Issue 1 (2017)


Abstract
This report documents the program and the outcomes of Dagstuhl Seminar 17042 "From Characters to Understanding Natural Language (C2NLU): Robust End-to-End Deep Learning for NLP". The seminar brought together researchers from different fields, including natural language processing, computational linguistics, deep learning and general machine learning. 31 participants from 22 academic and industrial institutions discussed advantages and challenges of using characters, i.e., "raw text", as input for deep learning models instead of language-specific tokens. Eight talks provided overviews of different topics, approaches and challenges in current natural language processing research. In five working groups, the participants discussed current natural language processing/understanding topics in the context of character-based modeling, namely, morphology, machine translation, representation learning, end-to-end systems and dialogue. In most of the discussions, the need for a more detailed model analysis was pointed out. Especially for character-based input, it is important to analyze what a deep learning model is able to learn about language - about tokens, morphology or syntax in general. For an efficient and effective understanding of language, it might furthermore be beneficial to share representations learned from multiple objectives to enable the models to focus on their specific understanding task instead of needing to learn syntactic regularities of language first. Therefore, benefits and challenges of transfer learning were an important topic of the working groups as well as of the panel discussion and the final plenary discussion.

Cite as

Phil Blunsom, Kyunghyun Cho, Chris Dyer, and Hinrich Schütze. From Characters to Understanding Natural Language (C2NLU): Robust End-to-End Deep Learning for NLP (Dagstuhl Seminar 17042). In Dagstuhl Reports, Volume 7, Issue 1, pp. 129-157, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2017)


Copy BibTex To Clipboard

@Article{blunsom_et_al:DagRep.7.1.129,
  author =	{Blunsom, Phil and Cho, Kyunghyun and Dyer, Chris and Sch\"{u}tze, Hinrich},
  title =	{{From Characters to Understanding Natural Language (C2NLU): Robust End-to-End Deep Learning for NLP (Dagstuhl Seminar 17042)}},
  pages =	{129--157},
  journal =	{Dagstuhl Reports},
  ISSN =	{2192-5283},
  year =	{2017},
  volume =	{7},
  number =	{1},
  editor =	{Blunsom, Phil and Cho, Kyunghyun and Dyer, Chris and Sch\"{u}tze, Hinrich},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops-dev.dagstuhl.de/entities/document/10.4230/DagRep.7.1.129},
  URN =		{urn:nbn:de:0030-drops-72489},
  doi =		{10.4230/DagRep.7.1.129},
  annote =	{Keywords: Natural Language Understanding, Artificial Intelligence, Deep Learning, Natural Language Processing, Representation Learning}
}
Document
Sampling in Potts Model on Sparse Random Graphs

Authors: Yitong Yin and Chihao Zhang

Published in: LIPIcs, Volume 60, Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques (APPROX/RANDOM 2016)


Abstract
We study the problem of sampling almost uniform proper q-colorings in sparse Erdos-Renyi random graphs G(n,d/n), a research initiated by Dyer, Flaxman, Frieze and Vigoda [Dyer et al., RANDOM STRUCT ALGOR, 2006]. We obtain a fully polynomial time almost uniform sampler (FPAUS) for the problem provided q>3d+4, improving the current best bound q>5.5d [Efthymiou, SODA, 2014]. Our sampling algorithm works for more generalized models and broader family of sparse graphs. It is an efficient sampler (in the same sense of FPAUS) for anti-ferromagnetic Potts model with activity 0<=b<1 on G(n,d/n) provided q>3(1-b)d+4. We further identify a family of sparse graphs to which all these results can be extended. This family of graphs is characterized by the notion of contraction function, which is a new measure of the average degree in graphs.

Cite as

Yitong Yin and Chihao Zhang. Sampling in Potts Model on Sparse Random Graphs. In Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques (APPROX/RANDOM 2016). Leibniz International Proceedings in Informatics (LIPIcs), Volume 60, pp. 47:1-47:22, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2016)


Copy BibTex To Clipboard

@InProceedings{yin_et_al:LIPIcs.APPROX-RANDOM.2016.47,
  author =	{Yin, Yitong and Zhang, Chihao},
  title =	{{Sampling in Potts Model on Sparse Random Graphs}},
  booktitle =	{Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques (APPROX/RANDOM 2016)},
  pages =	{47:1--47:22},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-018-7},
  ISSN =	{1868-8969},
  year =	{2016},
  volume =	{60},
  editor =	{Jansen, Klaus and Mathieu, Claire and Rolim, Jos\'{e} D. P. and Umans, Chris},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops-dev.dagstuhl.de/entities/document/10.4230/LIPIcs.APPROX-RANDOM.2016.47},
  URN =		{urn:nbn:de:0030-drops-66706},
  doi =		{10.4230/LIPIcs.APPROX-RANDOM.2016.47},
  annote =	{Keywords: Potts model, Sampling, Random Graph, Approximation Algorithm}
}
  • Refine by Author
  • 1 Blunsom, Phil
  • 1 Cho, Kyunghyun
  • 1 Dyer, Chris
  • 1 Schütze, Hinrich
  • 1 Yin, Yitong
  • Show More...

  • Refine by Classification

  • Refine by Keyword
  • 1 Approximation Algorithm
  • 1 Artificial Intelligence
  • 1 Deep Learning
  • 1 Natural Language Processing
  • 1 Natural Language Understanding
  • Show More...

  • Refine by Type
  • 2 document

  • Refine by Publication Year
  • 1 2016
  • 1 2017

Questions / Remarks / Feedback
X

Feedback for Dagstuhl Publishing


Thanks for your feedback!

Feedback submitted

Could not send message

Please try again later or send an E-mail