BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//https://techplay.jp//JP
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALDESC:ポンタス氏(UCL)と米田氏(豊田工業大学)による
 講演会
X-WR-CALNAME:ポンタス氏(UCL)と米田氏(豊田工業大学)による
 講演会
X-WR-TIMEZONE:Asia/Tokyo
BEGIN:VTIMEZONE
TZID:Asia/Tokyo
BEGIN:STANDARD
DTSTART:19700101T000000
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:JST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:695238@techplay.jp
SUMMARY:ポンタス氏(UCL)と米田氏(豊田工業大学)による講演
 会
DTSTART;TZID=Asia/Tokyo:20180914T123000
DTEND;TZID=Asia/Tokyo:20180914T140000
DTSTAMP:20260504T045730Z
CREATED:20180906T141145Z
DESCRIPTION:イベント詳細はこちら\nhttps://techplay.jp/event/69523
 8?utm_medium=referral&utm_source=ics&utm_campaign=ics\n\nAIP will host tw
 o talks at 12:30-14:00 on Friday\, September 14 at AIP (COREDO Nihonbashi
  15F). Everyone who received this email are welcome to attend.\nhttps://a
 ip.riken.jp/access/\n\n1) Pontus Stenetrop (Senior Research Associate at 
 University College London)\n "Recent Trends in Natural Language Processin
 g: Reading Comprehension and Fact Verification – with additional advice
  for young researchers”\n\n2) Takuma Yoneda (Master Student at Toyota I
 nstitute - intern at Pontus’s group at UCL)\n “Work of FEVER(Fact Ext
 raction and VERification) at UCL and Internship experience)\n\nABSTRACT\n
 1) Talk by Pontus Stenetrop\nNatural Language Processing\, along with all
  of Artificial Intelligence\, has been receiving considerable attention o
 ver the last few years. In this talk\, we want to highlight two very acti
 ve areas of research: Reading Comprehension and Fact Verification. Lastly
 \, we hope to give concrete advice based on our own experiences on how to
  navigate and make good research progress as a young researcher – we wi
 ll also be available after the talk for personal questions.\n\nReading Co
 mprehension is the task of answering questions for a given text. We will 
 give a historical background of the task\, the recently introduced datase
 ts\, and modelling approaches for solving the task. We will then relate t
 his to recent research from ourselves and others on weaknesses of existin
 g datasets and models\, such as “easy samples”\, inability to handle 
 “multi-step” reasoning – predicting future lines of research.\n\n
 “Fake News” has been the focus of a lot attention over the last few y
 ears\, motivating research on automated fact verification. In this task\,
  a system is given a claim and is tasked to not only answer if the claim 
 is true or not\, but also to provide the relevant evidence. Our group at 
 University College London recently participated in the FEVER shared task\
 , where we finished second among 24 teams. We will give an outline of the
  task and our own system for solving it\, with a focus on the process of 
 how one approaches successfully participating in a shared task where time
  is very limited compared to “standard” academic research.\n\n2) Talk
  by Takuma Yoneda\n\nFEVER work\n\nDuring my intern at UCLMR\, I worked o
 n the shared task called FEVER(Fact Extraction and VERification)\, which 
 is a new shared task showing up in the context of increasing need for aut
 omated fact-checking.\nGiven a claim\, this task requires a model to answ
 er whether the claim is true/false as well as providing the evidences bas
 ed on large corpus such as Wikipedia.\nI work on it with two researchers 
 in UCLMR and build up a model that is roughly composed of Information Ret
 rieval (IR) part\, which retrieves evidences for a claim\, and Natural La
 nguage Inference (NLI) part\, which predicts whether retrieved evidences 
 can support/refute the claim\, and an aggregation part\, which aggregates
  predictions for each claim-evidence pair.\nAs a result\, our model got t
 he 2nd place out of 24 groups/individuals in the shared task.\nI’ll tal
 k briefly about the shared-task\, architecture of the model and how I wor
 ked collaboratively with other (foreign) researchers.\n\nInternship exper
 ience\n\nWhile staying at one of the top-level laboratories in a foreign 
 country\,  I found that the atmosphere in the laboratory and students’ 
 attitude toward their research are totally different from those in Japan.
 \nStudents are so active and they are enjoying discussion on research top
 ic almost at every moment. They also find interesting papers/blogs every 
 day and share them on slack.\nThe intriguing things here is that even pro
 fessors are also involved in these activities\, which makes these interac
 tions more valuable and reasonable.\nIn addition to these\, we could have
  lots of opportunities to listen to and talk with famous researchers comi
 ng from other research institutes such as DeepMind\, Google and Microsoft
  since they gather in London.\nI’ll talk about my experience in the int
 ernship especially focusing on the difference in atmosphere/belief toward
  research compared with those in Japan\, and how much young researchers c
 ould get through the intern in a foreign country.
LOCATION:AIP （コレド日本１５階） 〒103-0027 東京都中央区
 日本橋1-4-1 日本橋一丁目三井ビルディング 15階
URL:https://techplay.jp/event/695238?utm_medium=referral&utm_source=ics&utm
 _campaign=ics
END:VEVENT
END:VCALENDAR