Difference between revisions of "Online Apertium Workshop 2020"

From Apertium
Jump to navigation Jump to search
(Start Online Apertium Workshop page)
 
 
(21 intermediate revisions by 6 users not shown)
Line 1: Line 1:
 
A two-session online workshop to discuss how information flows from one module to another in the Apertium pipeline.
=Online Apertium Workshop 2020=
 
   
  +
All Apertium developers welcome!
A two-session online workshop to discuss how information flows from one module to another in the Apertium pipeline,
 
   
  +
[http://meet.google.com/mff-qmap-zxb Google Meet link]
==Session #1: Tuesday, June 30, 1400Z–1600Z==
 
   
  +
Sessions will be moderated by Jonathan N. Washington.
There are only two presentations so far (only three of you have registered; 8 people have selected dates in the Doodle).1
 
  +
 
==Session #1: Tuesday, June 30, 1400–1600 UTC==
   
  +
(Under construction)
* 1400 Welcome and adjustments
 
* 1410 "Reading-bound data as inline secondary tags", Tino Didriksen
 
** "Reading-bound data is best transported as inline secondary tags, proven both by practical experience and theoretical complexity."
 
* 1425 "Moving all Multiwords to Apertium-separable", Daniel Swanson
 
**"I will describe what would be involved in migrating most or all multiwords from monodixes to language-specific -separable rules, including the extent to which it can be automated and how it might impact the other debates surrounding trimming and secondary information. (If this is deemed out-of-scope for this discussion, let me know, though I'm interested in attending regardless.)"
 
* 1440 ???
 
* 1455 ???
 
* 1510 Open discussion
 
   
  +
*10-minute/10-slide talks followed by 10 minute Q&A.
== Session #2: Thursday, July 2, 1400Z–1600Z==
 
 
** 1400 UTC Welcome and adjustments
  +
** 1410 UTC "Rule-based machine translation and the Apertium paradigm", Francis Tyers
  +
*** I will present a short overview of rule-based machine translation paradigms and situate Apertium within them.
  +
** 1425 UTC "Why we want to eliminate trimming and the motivation for secondary information", Tanmai Khanna
  +
***I will try to give a background about the entire secondary information discussion. This started as a project to eliminate trimming, and secondary information came out of it as a possible solution to that and other problems, such as markup handling. I will present the pros and cons of trimming and why we want to eliminate it, and the possible solutions.
 
** 1450 UTC "[https://docs.google.com/presentation/d/1LBcBs3KdzfS7vl6Sxe0UtOMLpWNMM6ciGS_YPCnxTr0 Reading-bound data as inline secondary tags]", Tino Didriksen
 
*** "Reading-bound data is best transported as inline secondary tags, proven both by practical experience and theoretical complexity."
 
** 1515 UTC "Moving all Multiwords to Apertium-separable", Daniel Swanson
 
***"I will describe what would be involved in migrating most or all multiwords from monodixes to language-specific -separable rules, including the extent to which it can be automated and how it might impact the other debates surrounding trimming and secondary information. (If this is deemed out-of-scope for this discussion, let me know, though I'm interested in attending regardless.)"
  +
* 1540 UTC Short open discussion
  +
* 1555 UTC Closing
   
 
== Session #2: Thursday, July 2, 1400 UTC–1600 UTC==
General discussion
 
  +
 
* 1400 UTC General discussion
  +
** [https://docs.google.com/presentation/d/1NEs_4fvP1M9VQd100tcJzGG_yEbypW59i85_JpBssNg/edit?usp=sharing <s>Weird</s> Non-Canonical things we do in Catalan-related pairs] -> Xavi Ivars (I may be late for the meeting)
  +
** Ugly hacks and where to find them [[/Ugly hacks]]
  +
*** I dont think Ill have slides/speech but Id like to collect some samples relevant for trimming / separable etc. everyone to just add to list from their experiences of frustration with other peoples monodixes etc., hopefully well have canonical solutions too! --[[User:TommiPirinen|Tommi Pirinen a.k.a. Flammie]] ([[User talk:TommiPirinen|talk]]) 09:09, 1 July 2020 (UTC)
  +
** [https://docs.google.com/presentation/d/1uJYVTCg8hw-90_XiE3fc3GWQRvdR-CcUgbToYOnAcQc Why we want to transport secondary information] - [[User:Tino Didriksen|Tino Didriksen]] ([[User talk:Tino Didriksen|talk]])
  +
** [https://docs.google.com/presentation/d/185i8rnQ0-54bX8rbvl7p636jRnEVi5NkfdAtvKIT6pY/edit?usp=sharing Anaphora Resolution in Apertium] [[User:Khannatanmai|Tanmai Khanna]] ([[User talk:Khannatanmai|talk]])
  +
* 1500 UTC Conclusions and proposal to the Apertium Community
  +
* 1555 UTC Closing
  +
* https://docs.google.com/presentation/d/1BS1U-2UTR5WPBSfP2zYH_2ioJaeZSw5WSlA8m6Bboko/edit?usp=sharing
  +
  +
== How to connect ==
  +
We will use [http://meet.google.com/mff-qmap-zxb Google Meet]. Please join with your microphone off when you are admitted.
  +
  +
Please *sign up [https://docs.google.com/forms/d/1tqs1bcWdBLDCRhZOLZ7T_xn5_DJdzxL2RrbMvi2RQm8 here]* before the workshop.

Latest revision as of 15:49, 2 July 2020

A two-session online workshop to discuss how information flows from one module to another in the Apertium pipeline.

All Apertium developers welcome!

Google Meet link

Sessions will be moderated by Jonathan N. Washington.

Session #1: Tuesday, June 30, 1400–1600 UTC[edit]

(Under construction)

  • 10-minute/10-slide talks followed by 10 minute Q&A.
    • 1400 UTC Welcome and adjustments
    • 1410 UTC "Rule-based machine translation and the Apertium paradigm", Francis Tyers
      • I will present a short overview of rule-based machine translation paradigms and situate Apertium within them.
    • 1425 UTC "Why we want to eliminate trimming and the motivation for secondary information", Tanmai Khanna
      • I will try to give a background about the entire secondary information discussion. This started as a project to eliminate trimming, and secondary information came out of it as a possible solution to that and other problems, such as markup handling. I will present the pros and cons of trimming and why we want to eliminate it, and the possible solutions.
    • 1450 UTC "Reading-bound data as inline secondary tags", Tino Didriksen
      • "Reading-bound data is best transported as inline secondary tags, proven both by practical experience and theoretical complexity."
    • 1515 UTC "Moving all Multiwords to Apertium-separable", Daniel Swanson
      • "I will describe what would be involved in migrating most or all multiwords from monodixes to language-specific -separable rules, including the extent to which it can be automated and how it might impact the other debates surrounding trimming and secondary information. (If this is deemed out-of-scope for this discussion, let me know, though I'm interested in attending regardless.)"
  • 1540 UTC Short open discussion
  • 1555 UTC Closing

Session #2: Thursday, July 2, 1400 UTC–1600 UTC[edit]

How to connect[edit]

We will use Google Meet. Please join with your microphone off when you are admitted.

Please *sign up here* before the workshop.