森羅2021-MLキックオフミーティング開催[5/25]と森羅2021-LinkJP資料公開のお知らせ

2 views
Skip to first unread message

Masako Nomoto

unread,
May 20, 2021, 1:05:42 AM5/20/21
to shinr...@googlegroups.com

森羅2019メーリングリストの皆様:

森羅2021実行委員の野本です。

1) 森羅2021-ML(多言語分類タスク)
 
 昨年に引き続き、30言語のWikipediaを約220種類の意味的カテゴリーに分類するタスク
 の参加者を募集しています。(本メールの後半 3)に参加募集を添付します。)

 5月25日17:00からキックオフミーティング(オンライン,無料)を行いますので是非ご参加
 下さい。
 * 招待講演では前回SHINRA2020-MLでトップのグループのLe-Hong Phuong先生(VNU
 University of Science)にご講演いただく予定です。
 * 参加登録とキックオフミーティングの詳細についてはこちらのページをご覧ください。
 * 今年のタスクでは昨年のシステム結果の利用が可能です。また、リーダーボードのみの
 参加も可能です。
 * Slackにも是非ご参加ください。[参加用リンク]

2) 森羅2021-LinkJP(日本語Wikipediaリンクタスク)

 日本語Wikipediaページで見出しの項目について書かれた情報(属性値、例えば小説家の
 ページなら作品名等)をWikipediaページに紐付ける新タスクも参加者を募集しています。

 * 4月22日に開催したキックオフミーティングのビデオと資料をホームページで公開してい
 ますので、ぜひご覧ください。
 * Slackにもぜひご参加ください。[参加リンク]

3) 森羅2021-ML 参加募集
========================================================
                                  CALL FOR TASK PARTICIPATION

     SHINRA2021-ML: Categorization task of 30 language Wikipedias into fine-grained NE
     http://shinra-project.info/shinra2021ml/?lang=en

Data release: March 2021
Kick-off meeting: May 25, 2021 17pm(JST) *    
     Registration at https://c5dc59ed978213830355fc8978.doorkeeper.jp/events/121360
Result submission deadline: October 15, 2021
Final Report Meeting: December 2021

* 9am in London, 10am in Paris, 15pm in Hanoi, 16pm in Singapore, 18pm in Sydney, 1am
in San Francisco, 4am in New York; all on May 25.
=========================================================
 
*OVERVIEW

SHINRA is a resource creation project started in the year 2017, aiming to structure the
knowledge in Wikipedia. SHINRA2021-ML is a shared-task of text categorization, tackling
the problem of classifying 30-language Wikipedia entities in fine-grained categories
(Extended Named Entity).

The participants are expected to select one or more target languages. For each language,
using the categorized Wikipedia pages created from the categorized Japanese pages and
language-link information as the training data, run the system to classify the remaining
pages which are not linked from the Japanese pages.

After the task is over, the system outputs will be made public, and we (including the
participants) can combine the results to make better results (i.e. by Ensemble learning).
All those resources will be made public. We call this scheme "Resource by Collaborative
Contribution (RbCC)" and we are expecting many participants with a good will.


*HOW TO PARTICIPATE

Please go to the SHINRA2021-ML task HP (http://shinra-project.info/shinra2021ml/?lang=en),
and follow "Participation procedures".
Note for the participants of SHINRA2020-ML task, unlike last year, SHINRA2021-ML will
run by ourselves, not by NTCIR, so no registration at NTCIR is needed.


*CONTACT

Email to the organizers:
shinra20...@googlegroups.com

SHINRA2021-ML homepage:
http://shinra-project.info/shinra2021ml/?lang=en
Reply all
Reply to author
Forward
0 new messages