NTCIR STC

NTCIR Short Text Conversation Task

[NTCIR STC @ntcirstc]

Yes @ntcirdialeval is my daughter. As I'm retiring, please follow her instead!

Good luck my dear little girl!

[NTCIR DialEval @ntcirdialeval]

Hi, I'm the NTCIR Dialogue Evaluation Task. I have been accepted for #ntcir15 .
My mother is @ntcirstc . I have two subtasks: Dialogue Quality (DQ) and ND (Nugget Detection). These subtasks are the same as the #ntcir14 STC-3 DQ and ND subtasks.

[NTCIR STC @ntcirstc]

Draft overview + official results for the Dialogue Quality (DQ) and Nugget Detection (ND) subtasks have been sent to the participants!

https://t.co/qUIJ2OYD9Z

[NTCIR STC @ntcirstc]

The training and test data (in Chinese and English) for
DQ (Dialogue Quality) and ND (Nugget Detection) have just been released!
https://t.co/t2vcw8BXpH

The English data are manual translations of the original Chinese Helpdesk-Customer dialogues on Weibo.

[NTCIR STC @ntcirstc]

We already have 14 registered teams but we have decided to extend the task registration deadline to Sunday 9th Steptember!

Please register! Have fun with our new tasks!

https://t.co/OFfXzBJaWc

[NTCIR STC @ntcirstc]

Task registration for #ntcir14 STC-3 will close tomorrow so register NOW if you haven't!
https://t.co/WyIPaqYK5R

Task details: https://t.co/qUIJ2OYD9Z

[NTCIR STC @ntcirstc]

Please participate in the #ntcir14 Short Text Conversation task (STC-3) if you are interested in dialogue systems and/or dialogue evaluation!

Slides (updated June 6):
https://t.co/gYdsEcUip3

[NTCIR STC @ntcirstc]

DQ: given a customer-helpdesk dialogue, return an estimated distribution over user ratings e.g. customer satisfaction.

ND: given a dialogue, for each customer or helpdesk utterance block, return an estimated distribution over utterance types, e.g. trigger, goal, not-a-nugget.

[NTCIR STC @ntcirstc]

We have updated the slides for the Dialogue Quality subtask and the Nugget Detection subtask (May 2). Please visit our website!
https://t.co/qUIJ2OYD9Z
#ntcir14

[Tetsuya Sakai (酒井哲也) @tetsuyasakai]

#sigir2018 short paper accepted: Comparing Two Binned Probability Distributions for Information Access Evaluation

The measures will be used in the #ntcir14 @ntcirstc task.

[NTCIR STC @ntcirstc]

Short Text Conversation, THE largest task of NTCIR, is back!
Check out the tentative info of the Chinese Emotional Conversation Generation, Dialogue Quality, and Nugget Detection subtasks at https://t.co/qUIJ2OYD9Z
#ntcir14

[NTCIR STC @ntcirstc]

Please visit the #ntcir14 @ntcirstc (STC-3) poster at #ntcir13!

[NTCIR STC @ntcirstc]

I have been accepted as an #ntcir14 task! Many thanks to the PC.

[NTCIR STC @ntcirstc]

64 runs are retrieval-based just like STC1, while 56 runs are generation-based.
As many as 15 teams tried the generation-based approach! https://t.co/v5emh8Ubkn

[NTCIR STC @ntcirstc]

Sorry, correction - 120 Chinese runs from 22 teams.

[NTCIR STC @ntcirstc]

Sorry, correction - 120 Chinese runs from 22 teams.

[Yuki Arase (荒瀬 由紀) @Yuki_arase]

NTCIR Short Text Conversation (STC2) 日本語タスクの詳細が公開されました。
データ配布も開始されております。参加登録は7/23までです。
皆様のご参加お待ちしています！
https://t.co/cRpyBHyirI

[Yuki Arase (荒瀬 由紀) @Yuki_arase]

NTCIR13 STC2 (Japanese) web site is now open!
Registration deadline: 23 July
https://t.co/cRpyBHyirI

[NTCIR STC @ntcirstc]

In the end, we received 141 Chinese runs from 22 teams.
Not bad, not bad at all!
#ntcir13

[NTCIR STC @ntcirstc]

We have 37 participating groups now and we are still the largest #ntcir13 task! Task registration for STC-2 will be open until April 30.

[NTCIR STC @ntcirstc]

We have 34 participating groups now and we are still the largest #ntcir13 task! Task registration for STC-2 will be open until April 30.

[NTCIR STC @ntcirstc]

STC-2 evaluation criteria and run submission instructions:
https://t.co/bJqCpIxcqv