Typecraft v2.5
Jump to: navigation, search

Difference between revisions of "Classroom:Norwegian Grammar Checking"

(Using the grammar checker relative to a systematic course in Norwegian)
(A sentence construction game)
 
(136 intermediate revisions by 3 users not shown)
Line 1: Line 1:
====Norwegian Grammar Checking====
+
[[Korektor gramatyczny dla języka norweskiego]]    ([[Ćwiczenia z gramatyki norweskiej z polskim komentarzem]])
  
On clicking on the icon below (the picture of a troll), you get to an automatic parser for Norwegian designed to tell you, in case a sentence contains a grammatical mistake, what the mistake is.
+
[[Controllo della grammatica norvegese]];  [[Messaggi di feedback]]
  
 +
[[Feedback messages]]
  
 +
[[Rückmeldungen]]
  
[[File:Troll1.jpeg‎|100px|link=http://daria.idi.ntnu.no:8080/norsource/parse ]]
+
[[ 挪威语语法检测软件]]
  
  
Please click on the icon. In the window that comes up, you feed a sentence into the parser by writing or copying it into the open line, and pushing the button 'Analyze'.
 
  
For instance, consider the following sentence:
+
===Description===
  
''Jeg liker du.''
+
This device is designed to provide grammaticality feedback for sentences of your choice.
  
In this sentence, "du" is used as an object, but the form "du" can only be used as a subject. If you send the sentence through the parser, this is also what the parser will tell you. Thus, glue the sentence into the parse line, and push 'Analyze'. You get the following message:
 
  
"The word "du" is marked as the wrong case, try using "deg" instead".  
+
<span style="color:black">Please click on the Troll picture</span> [[File:Troll1.jpeg‎|border|90px|link=http://regdili.hf.ntnu.no/studentAce/parse]] <span style="color:black">to get to the Grammar Checker.</span>
  
To the left below this message, there is a small square with a number inside. Click on this square, and you get a new window showing, as 'generated sentence', the recommended form ''Jeg liker deg''.
 
  
You can in principle use sentences of any length, but for the purpose of getting rapid feedback and precise advice about what (if anything) is wrong, we recommend using short sentences.
 
  
 +
In the window that comes up, write one or more sentences (up to ten sentences at the time, with line-shift between, and each sentence no more than 10 words) into the open mask, and push the button 'Analyze'.
  
===Using the grammar checker relative to a systematic course in Norwegian===
+
If a string is grammatically wellformed, you will get the response:
  
A course like '''NoW''' given at NTNU has all its course material online, and the parser can be used in synchrony with the lessons of the course. The parser will then use the vocabulary of the course, and have all the grammar rules that the course provides. The parser allows you to test freely chosen sentences that you write.
+
          ''A wellformed Norwegian sentence''
  
One way of proceeding is that you refer to any sample text used in the course, and on a prompt question related to the text, write a short story/text as answer, using mainly the words used in the text, but in any grammatical structuring that you want to test out. For instance, the text may be the one for Lesson 5 in NoW, called "Maria" and illustrating inflections for Past tense, rendered below:
 
  
 +
If the string is ''not'' grammatically wellformed, there are two possibilities:
  
  
[[File:Bakklandet.jpg‎|thumb|right| Bakkland kafé]]
+
'''Error diagnosis'''
  
===5 MARIA===
+
The first possibility is that the system tells you ''what'' is wrong about the string. For instance, for the ungrammatical string
  
Kjære mamma og pappa!
+
          ''Jeg liker du.''
  
I dag '''syklet''' jeg til Bakklandet, et koselig område i nærheten av Trondheim sentrum. På Bakklandet er det mange gamle trehus i forskjellige farger. Det er også mange hyggelige kaféer og fine restauranter der. Jeg '''møtte''' Anna på en av kaféene, Dromedar Kaffebar. Jeg '''bestilte''' et stort stykke sjokoladekake og en kopp kaffe. Jeg '''betalte''' over 80 kroner for kakestykket og kaffen. Alt er så dyrt i Norge! Men kaka '''var''' kjempegod. Jeg '''spurte''' servitøren om oppskrifta, men den '''var''' hemmelig, sa han.
+
where the subject form "du" is used as an object, you will get the response:
  
Anna er ei veldig hyggelig jente. Jeg er så glad for at vi er naboer. Vi '''satt''' på kaféen i over to timer, for vi '''hadde''' så mye å snakke om. Og tenk – sykkelen '''stod''' utenfor hele tiden – uten lås! Jeg '''glemte''' å låse den da jeg '''gikk''' inn. Jeg '''tror''' at Trondheim er en trygg by, og det er få fattige mennesker her.
+
      ''The word "du" is marked with the wrong case, try using "deg" instead.''  
  
[[File:Trampe.jpeg‎|thumb|left|sykkelheis i Bakklandet]]
+
Accompanying such an error message, there may appear a button 'Generate'. By clicking it, you can get a new window showing a recommended version of the intended sentence, in the present case:
Nordmenn er rare. De er snille og hyggelige, men de holder ikke døra åpen for deg. I går '''fikk''' jeg døra på postkontoret i ansiktet, og i dag døra på kaféen.
+
  
Pass godt på hverandre, og hils bestefar fra meg.
+
          ''Jeg liker deg.''
  
Hilsen Maria
+
(To write a new sentence for feedback, return to the previous window.)
  
 +
Next to the 'Generate' button, there is also a button saying ''More description'', which takes you to a brief description of the relevant part of Norwegian grammar.
  
  
 +
'''Ungrammaticality report'''
  
 +
The second possibility is that the grammar recognizes a string as ungrammatical, but no specific message has been designed for the error type. The interface will then simply report:
  
 +
          ''Ungrammatical in Norwegian.''
  
  
 +
'''Some constraints'''
  
 +
The sentences entered should always have a verb. Apart from imperatives, sentences should also have a subject.  Moreover, ''interjections'' (like "yes", "hi!", and so) are not accepted.
  
 +
If the system does not contain a certain word used in the input string, the response will inform you that this is so:
  
 +
            ''Lexicon entry `...' is missing''
  
 +
Please note that the system does not have many non-Norwegian proper names in its repository, so that writing ''Jeg heter ...'' with a non-Norwegian proper name in the dotted area may well result in the above message.
  
 +
===Phenomena for which you can get feedback, and possible feedback messages===
  
Suppose that the prompt question is
 
  
 +
As a grammar checker, the system is not concerned with word spelling, and thus does not include what is normally referred to as a ''spellchecker''. (an example of a pilot system checking just spelling of words and not syntax can be seen at [[Spellchecker for Luganda - Pilot]]). Also, among the millions of ways in which one can make a grammatical mistake in a language, only a few are covered here in terms of specific feedback messages (other than just "Ungrammatical in Norwegian").
 +
For an overview of phenomena for which you can get such feedback messages, see [[Feedback messages]].
  
''Write 4-5 sentences in answer to the question:
 
"Hva gjorde du og Anna?"''
 
  
 +
For some background for the system, see [[A Norwegian Grammar Sparrer]].
  
and that you write the following:
+
===A sentence construction game===
  
  
''Anna og jeg syklet til en kafe.  Hun er min naboen. Vi snakte i to timer. Ingen stjelte syklen min, selv om den ikke var låst.''
+
For the new game '''Norwegian Grammar Scrabble''', go to
  
 +
see [[Score points for Norwegian sentences!]]
  
====What is wrong here?====
+
and [https://regdili.hf.ntnu.no/scrabble/page/#/start '''''web demo'''''].
  
You can copy and paste one sentence after the other into the Grammar checker.
+
===Background system===
 +
For background on the computational grammar running the system, see [[Norwegian HPSG grammar NorSource]].
  
If you write the sentences first in a normal text editor, spell-checkers built into the editor may detect some of the mistakes, typically wrong spellings of words and wrong inflections. You may use this as a first sift, and then see what the parser says about the grammatical structure.
 
  
Relative to each phenomenon for which the parser gives an error message, TypeCraft will have a page with information about the phenomenon involved, in the form of short definitions, examples of correct forms for the phenomenon, and, for further investigation, more detailed explanations and reference to literature, as well as reference to other languages manifesting the same or similar phenomena. It of course also has a reference to the lesson in NoW where the phenomenon is explained.
 
  
  
For instance, for errors in the use of Past tense, in Norwegian called 'Preteritum', a comment like the following will be supplied:
+
--[[User:Lars Hellan|Lars Hellan]] 15:11, 14 April 2014 (UTC)
  
 +
[[Category:Norwegian Grammar]]
  
<span style="color:green">GRAMMAR RULES</span>
 
  
===The past tense (preteritum)===
+
[http://www4.clustrmaps.com/user/8abdaf33 http://www4.clustrmaps.com/stats/maps-no_clusters/www.typecraft.org-thumb.jpg]
 
+
The past tense (preteritum) refers to something that happened at a specific time in the past, or happened repeatedly in the past.
+
 
+
Norwegian verbs can be classified as either WEAK (endings are added to the root of the verb to form the past tense) or STRONG (involving a vowel change in the root -- often no ending at all.)
+
+
====Basic Grammar Rules====
+
There are fewer strong verbs than weak verbs, so you should learn to recognise those and then assume that the rest are weak.
+
 
+
Memorise the strong verb. The rest of the verbs are then the weak ones.
+
 
+
Also for the weak verb it must be learned how to form the past form from the verb root. However, the past tense of weak verbs follows an easily recognisable pattern.
+
 
+
There are four different endings for weak verbs (-et | -te | de | dde). These endings are always added to the stem or root of the verb. This root is found by dropping the unstressed -e on the end of the infinitive, if there is one.
+
 
+
====<span style="color:green"> Exercises</span>====
+
'''To do an exercise click on the rat!  :)'''
+
 
+
[[File:Exercise.jpeg‎|100px|link=http://typecraft.org/TCEditor/1425/ ]]
+

Latest revision as of 19:17, 6 March 2024

Korektor gramatyczny dla języka norweskiego (Ćwiczenia z gramatyki norweskiej z polskim komentarzem)

Controllo della grammatica norvegese; Messaggi di feedback

Feedback messages

Rückmeldungen

挪威语语法检测软件


Description

This device is designed to provide grammaticality feedback for sentences of your choice.


Please click on the Troll picture Troll1.jpeg to get to the Grammar Checker.


In the window that comes up, write one or more sentences (up to ten sentences at the time, with line-shift between, and each sentence no more than 10 words) into the open mask, and push the button 'Analyze'.

If a string is grammatically wellformed, you will get the response:

          A wellformed Norwegian sentence


If the string is not grammatically wellformed, there are two possibilities:


Error diagnosis

The first possibility is that the system tells you what is wrong about the string. For instance, for the ungrammatical string

          Jeg liker du.

where the subject form "du" is used as an object, you will get the response:

      The word "du" is marked with the wrong case, try using "deg" instead. 

Accompanying such an error message, there may appear a button 'Generate'. By clicking it, you can get a new window showing a recommended version of the intended sentence, in the present case:

         Jeg liker deg.

(To write a new sentence for feedback, return to the previous window.)

Next to the 'Generate' button, there is also a button saying More description, which takes you to a brief description of the relevant part of Norwegian grammar.


Ungrammaticality report

The second possibility is that the grammar recognizes a string as ungrammatical, but no specific message has been designed for the error type. The interface will then simply report:

          Ungrammatical in Norwegian.


Some constraints

The sentences entered should always have a verb. Apart from imperatives, sentences should also have a subject. Moreover, interjections (like "yes", "hi!", and so) are not accepted.

If the system does not contain a certain word used in the input string, the response will inform you that this is so:

            Lexicon entry `...' is missing

Please note that the system does not have many non-Norwegian proper names in its repository, so that writing Jeg heter ... with a non-Norwegian proper name in the dotted area may well result in the above message.

Phenomena for which you can get feedback, and possible feedback messages

As a grammar checker, the system is not concerned with word spelling, and thus does not include what is normally referred to as a spellchecker. (an example of a pilot system checking just spelling of words and not syntax can be seen at Spellchecker for Luganda - Pilot). Also, among the millions of ways in which one can make a grammatical mistake in a language, only a few are covered here in terms of specific feedback messages (other than just "Ungrammatical in Norwegian"). For an overview of phenomena for which you can get such feedback messages, see Feedback messages.


For some background for the system, see A Norwegian Grammar Sparrer.

A sentence construction game

For the new game Norwegian Grammar Scrabble, go to

see Score points for Norwegian sentences!

and web demo.

Background system

For background on the computational grammar running the system, see Norwegian HPSG grammar NorSource.



--Lars Hellan 15:11, 14 April 2014 (UTC)


www.typecraft.org-thumb.jpg