générer un nouvel exercice genre cliquer sur des mots

[alexiye]

Décrivez la solution que vous aimeriez trouver dans Chamilo
Créer une nouvel exercice dans lequel on pourrait cliquer sur des mots.
Contexte additionnel
Donnez-nous (si vous en avez) des éléments additionnels qui pourraient nous aider dans l'implémentation ou l'identification d'une solution.
exemple (voir l'exercice en fin de page):
http://www.scalpa.info/francais/grammaire/la_fonction_des_mots/le_sujet_et_le_verbe/index.php

[ywarnier]

Bonjour @alexiye
C'est une bonne idée mais ça représente une bonne quantité de travail pour un résultat qui est assez spécifique (qui ne touche pas un public très large). Pas vraiment une priorité pour l'instant, mais la proposition est intéressante.

Je traduis donc pour une future implémentation:

This type of question would offer the possibility to distinguish words or terms in a text between different groups. So some terms are "marked" and match a specific group (in this example: subject, verb and an option to delete some terms, that we don't really want here).

The teacher first gives the groups (probably a text field with the groups names separated by ";"), then types a long text.
Then the teacher selects the first group and marks all the terms that are from this group.
Then the teacher selects the second group and marks all the terms that are from this group, and so on.
The teacher saves. The score is based on the total number of marked terms. One term cannot be part (for now) of two groups (because that would introduce a display style challenge).
The teacher can also "unmark" a certain term (he just selects it again).
The teacher gives a total max score for the question.
Somehow the text must be marked already, separating terms with a "span" with an "id" that represents the sequence identifier of that term in the text. Every time the question is saved, this has to be re-generated, because even one letter change in the text can break the whole thing.

Given the fact that the same term can appear twice, we need to make sure we number the terms (maybe get a count of characters in the text and just say that from character 12 to 19, the term is part of group of

The student starts the question and selects the group, then marks one or several terms, then selects another group, and so on, then he saves.

We then generate an array with the terms marked for each group, and compare it to the array of terms marked by the student. The number of terms in the teachers array is counted and kept as the maximum. Every match between the teacher array and the student array is considered a point earned. The total count of right answers is divided by the max score, and that gives us a percentage for the score to give to the student.

At the design level, the text must be typed in without special HTML style. We need raw text, so some option should allow us to insert it without style.
Then a process must go through the text and mark each word with a span. There's something of a headache here to find out how to select two words that go together.
Then the teacher can select the terms he wants to assign to a group (which exclude terms already assigned to another group) and that is saved as a serialized array.
Whenever the teacher assigns a group to a term, that term is underlined (visually) with the color of the group.

There's a lot of work in the design of the user interface here...

[alexiye]

Merci pour l'intérêt et le temps que vous avez apporté à ma requête. Je travaille depuis six ans avec chamilo dans ma classe de Ce1-Ce2 et votre LMS est un outil merveilleux pour les élèves. Cela s'avère très utile au niveau de la différenciation des apprentissages et une aide capitale pour les apprenants en difficulté. ils adorent valider les compétences et la roue qui verdit les motive. Bravo à votre équipe et merci. Alexandre Bigan

…

Le 19 déc. 2018 à 17:28, Yannick Warnier ***@***.***> a écrit : Bonjour @alexiye C'est une bonne idée mais ça représente une bonne quantité de travail pour un résultat qui est assez spécifique (qui ne touche pas un public très large). Pas vraiment une priorité pour l'instant, mais la proposition est intéressante. Je traduis donc pour une future implémentation: This type of question would offer the possibility to distinguish words or terms in a text between different groups. So some terms are "marked" and match a specific group (in this example: subject, verb and an option to delete some terms, that we don't really want here). The teacher first gives the groups (probably a text field with the groups names separated by ";"), then types a long text. Then the teacher selects the first group and marks all the terms that are from this group. Then the teacher selects the second group and marks all the terms that are from this group, and so on. The teacher saves. The score is based on the total number of marked terms. One term cannot be part (for now) of two groups (because that would introduce a display style challenge). The teacher can also "unmark" a certain term (he just selects it again). The teacher gives a total max score for the question. Somehow the text must be marked already, separating terms with a "span" with an "id" that represents the sequence identifier of that term in the text. Every time the question is saved, this has to be re-generated, because even one letter change in the text can break the whole thing. Given the fact that the same term can appear twice, we need to make sure we number the terms (maybe get a count of characters in the text and just say that from character 12 to 19, the term is part of group of The student starts the question and selects the group, then marks one or several terms, then selects another group, and so on, then he saves. We then generate an array with the terms marked for each group, and compare it to the array of terms marked by the student. The number of terms in the teachers array is counted and kept as the maximum. Every match between the teacher array and the student array is considered a point earned. The total count of right answers is divided by the max score, and that gives us a percentage for the score to give to the student. At the design level, the text must be typed in without special HTML style. We need raw text, so some option should allow us to insert it without style. Then a process must go through the text and mark each word with a span. There's something of a headache here to find out how to select two words that go together. Then the teacher can select the terms he wants to assign to a group (which exclude terms already assigned to another group) and that is saved as a serialized array. Whenever the teacher assigns a group to a term, that term is underlined (visually) with the color of the group. There's a lot of work in the design of the user interface here... — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

[andreboivin]

Une de mes élèves enseignante a trouvé une solution originale. Elle utilise l'outil zone sur image. Elle converti son texte en image et applique des zones qui correspondent aux questions.
Une solution de rechange en attendant...