Task ideas for Google Code-in - Revision history

Firespeaker: /* Task ideas */

2019-12-05T06:23:22Z

Task ideas

Firespeaker: /* Task ideas */

2019-12-05T05:29:01Z

Task ideas

Firespeaker: /* Task ideas */

2019-12-05T05:17:11Z

Task ideas

Firespeaker: /* Task ideas */

2019-12-05T05:16:04Z

Task ideas

Firespeaker: /* Task ideas */

2019-12-05T05:01:17Z

Task ideas

Firespeaker: /* Task ideas */

2019-12-05T05:00:32Z

Task ideas

Firespeaker: /* Task ideas */

2019-12-05T04:46:26Z

Task ideas

Firespeaker: /* Task ideas */

2019-11-29T17:19:00Z

Task ideas

Firespeaker: /* Task ideas */

2019-11-29T17:17:19Z

Task ideas

Firespeaker: /* Task ideas */

2019-11-29T17:16:06Z

Task ideas

@@ Line 377: / Line 377: @@
  |beginner=no
 }}
-<!-- {{Taskidea
- |type=quality
- |mentors=*
- |tags=localisation
- |title=Complete website localisation in a language not fully localised
- |description=
- |multi=
-}} -->
-<!-- NEW TASKS BELOW -->
@@ Line 442: / Line 431: @@
  |dup=10
 }}
+<!-- {{Taskidea
+ |type=quality
+ |mentors=*
+ |tags=localisation
+ |title=Complete website localisation in a language not fully localised
+ |description=
+ |multi=
+}} -->
+<!-- NEW TASKS BELOW -->
 </table>

@@ Line 395: / Line 395: @@
  |tags=video, tutorial
  |title=Video tutorial: installing Apertium, adding to dictionary, and submitting a PR
- |description=Post a video online that (1) demonstrates how to install Apertium on an operating system of your choice, (2) demonstrates how to clone and compile an Apertium translation pair of your choice, (3) shows how to add a new word to the dictionary (categorised correctly), and (4) shows how to submit the updated dictionary as a pull request to Apertium's git repository.  Add a link to the video on the [http://wiki.apertium.org/wiki/Installation#Installation_Videos installation videos page] of the Apertium wiki.<br/>The title of the video should make it easy to find, and so should probably be similar to the title of this task.  We recommend a screencast with voice-over posted to YouTube, but the format and venue are up to you as long as it is publicly accessible for long term.  Here are [https://www.youtube.com/playlist?list=PLHldb9r6QkVFsuxlAoVS-OL32aurUOZLC some example videos] that are relevant but that could probably be improved upon.<br/>The video *does not have to be in English*; we can evaluate it in any of the following languages: %ZZZ%.  Please let us know when you claim the task what language you plan to create the video in, so that we know which mentor(s) should primarily work to evaluate your task.
+ |description=Post a video online that (1) demonstrates how to install Apertium on an operating system of your choice, (2) demonstrates how to clone and compile an Apertium translation pair of your choice, (3) shows how to add a new word to the dictionary (categorised correctly), and (4) shows how to submit the updated dictionary as a pull request to Apertium's git repository.  Add a link to the video on the [http://wiki.apertium.org/wiki/Installation#Installation_Videos installation videos page] of the Apertium wiki.<br/>The title of the video should make it easy to find, and so should probably be similar to the title of this task.  We recommend a screencast with voice-over posted to YouTube, but the format and venue are up to you as long as it is publicly accessible for long term.  Here are [https://www.youtube.com/playlist?list=PLHldb9r6QkVFsuxlAoVS-OL32aurUOZLC some example videos] that are relevant but that could probably be improved upon.<br/>The video **does not have to be in English**; we can evaluate it in any of the following languages: %ZZZ%.  Please let us know when you claim the task what language you plan to create the video in, so that we know which mentor(s) should primarily work to evaluate your task.
  |multi=200
  |beginner=yes

@@ Line 438: / Line 438: @@
  |tags=evaluation
  |title=Evaluate an existing Apertium translation pair on a text
- |description= Pick an existing Apertium language pair and get a parallel text for that language pair. Translate the text using the Apertium translation pair and evaluate the translation using an automatic evaluation metric like BLEU and/or evaluate it manually.<br/>The language pairs we can mentor for this task are the following: %ALLPAIRS%.<br/>[http://wiki.apertium.org/wiki/Task_ideas_for_Google_Code-in/Evaluation_of_translation_of_an_existing_pair Read more]...
+ |description= Pick an existing Apertium language pair and get a parallel text for that language pair. Translate the text using the Apertium translation pair and evaluate the translation using an automatic evaluation metric like BLEU and/or evaluate it manually.<br/>The language pairs we can mentor for this task (pending their existence) are the following: %ALLPAIRS%.<br/>[http://wiki.apertium.org/wiki/Task_ideas_for_Google_Code-in/Evaluation_of_translation_of_an_existing_pair Read more]...
  |multi=40
  |dup=10

@@ Line 427: / Line 427: @@
  |tags=xml, dictionaries
  |title=Post-edit 500 sentences of a public domain text
- |description=Many of our systems benefit from statistical methods used with (ideally public domain) bilingual data.  For this task, you need to translate a public domain text using an available machine translation system (Apertium preferred) and clean up the translations yourself manually.  Commit the post-edited texts (in plain text format) to an existing (via pull request) or if needed new github repository for the language pair in dev/ or texts/ folder.  The texts are subject to mentor approval.<br/>The language pairs we can mentor for this task are the following: %ALLPAIRS%.
+ |description=Many of our systems benefit from statistical methods used with (ideally public domain) bilingual data.  For this task, you need to translate a public domain text using an available machine translation system (Apertium preferred) and clean up the translation yourself manually.  Commit the source text and post-edited translation (in plain text format) to an existing (via pull request) or if needed new github repository for the language pair in dev/ or texts/ folder.  The texts are subject to mentor approval.<br/>The language pairs we can hypothetically mentor for this task (pending their existence) are the following: %ALLPAIRS%.
  |multi=40
  |dup=10

@@ Line 407: / Line 407: @@
  |description=Our translation systems require large lexicons so as to provide production-quality coverage of any input data.  This task requires the student to add 100 new words to a bidirectional dictionary.  Choose one of the language pairs listed below, and with the help of your mentor, identify some text in one of the two languages, and run the text through Apertium's translator for that language pair to identify 100 unknown forms.  As needed, add the stems of these forms to the individual languages' analysers in an appropriate way so that these words are analysed correctly.  Your submission should be in the form of a pull request to each of the appropriate repositories on GitHub.<br/>The language pairs we can mentor for this task are the following: %ALLPAIRS%.<br/>  [http://wiki.apertium.org/wiki/Task_ideas_for_Google_Code-in/Grow_bilingual More instructions for this task here]...
  |multi=40
- |dups=10
+ |dup=10
  |beginner=yes
 }}
@@ Line 418: / Line 418: @@
  |description=Our translation systems require large lexicons so as to provide production-quality coverage of any input data.  This task requires the student to add 250 new words to a bidirectional dictionary.  Choose one of the language pairs listed below, and with the help of your mentor, identify some text in one of the two languages, and run the text through Apertium's translator for that language pair to identify 250 unknown forms.  As needed, add the stems of these forms to the individual languages' analysers in an appropriate way so that these words are analysed correctly.  Your submission should be in the form of a pull request to each of the appropriate repositories on GitHub.<br/>The language pairs we can mentor for this task are the following: %ALLPAIRS%.<br/>  [http://wiki.apertium.org/wiki/Task_ideas_for_Google_Code-in/Grow_bilingual More instructions for this task here]...
  |multi=40
- |dups=10
+ |dup=10
  |beginner=no
 }}
@@ Line 429: / Line 429: @@
  |description=Many of our systems benefit from statistical methods used with (ideally public domain) bilingual data.  For this task, you need to translate a public domain text using an available machine translation system (Apertium preferred) and clean up the translations yourself manually.  Commit the post-edited texts (in plain text format) to an existing (via pull request) or if needed new github repository for the language pair in dev/ or texts/ folder.  The texts are subject to mentor approval.<br/>The language pairs we can mentor for this task are the following: %ALLPAIRS%.
  |multi=40
- |dups=10
+ |dup=10
  |beginner=yes
 }}
@@ Line 440: / Line 440: @@
  |description= Pick an existing Apertium language pair and get a parallel text for that language pair. Translate the text using the Apertium translation pair and evaluate the translation using an automatic evaluation metric like BLEU and/or evaluate it manually.<br/>The language pairs we can mentor for this task are the following: %ALLPAIRS%.<br/>[http://wiki.apertium.org/wiki/Task_ideas_for_Google_Code-in/Evaluation_of_translation_of_an_existing_pair Read more]...
  |multi=40
- |dups=10
+ |dup=10
 }}