When you frequently use a certain tier type with a limited number of annotation values, it might be a good idea to associate a Controlled Vocabulary (CV) with it. Such a CV consists of a number of predefined values that a user can choose from when editing an annotation, thus making the task of the annotator less error-prone. E.g. one can choose to create a controlled vocabulary for part of speech tagging, as the tags that are used often reoccur. In order to do this select Edit > Edit Controlled Vocabularies. The following dialog will appear:
Figure 2.33. Edit controlled vocabulary
The undo function in the CV dialog window only works as long as that window is active. Once it is closed changes cannot be undone any more.
Also see the notes concerning Associating an annotation with an external resource.
By clicking More Options... (not yet shown in the figure above) you can choose a color that will fill the lower part of every annotation frame containing the selected CV entry. Moreover, you can choose a shortcut key to edit an annotation with a single key stroke.
The color chooser has four tabs. The last three contain different ways to choose a color, which is subsequently displayed in the lower part of the window. In the first tab you can add or insert the color displayed below and you can copy, paste and delete the selected color. The list of favourite colors is saved and used the next time you start ELAN.
New in version 4.7.0 is the option to set one or more languages for a CV. This allows you to have one CV that contains a certain vocabulary in both English and German, for instance. You can choose from nearly 7700 ISO specified languages.
This change in 4.7.0 introduced an incompatibility with earlier versions of ELAN;
the structure of the .eaf
changed to some extent. As a result,
when opening a new .eaf
in an older ELAN version, the entries of
CV's will be missing!
By default, the language for a CV is set to 'undetermined (und)'. To change this, click the drop-down menu and select Edit Languages...
Figure 2.34. Edit CV Languages
From the dialog that appears, select the desired language and click Change. The undetermined language in the upper drop-down menu will now be replaced with the chosen language. see Defining languages for your content to edit the list of available languages.
Figure 2.35. Edit CV Languages 2
If you would like to add an additional language, select the desired language from the lower drop-down menu, and click Add afterwards. The chosen language will be added to the CV languages, and will be visible in the upper drop-down menu. When you are done adding or changing languages, click Close to close the dialog.
Figure 2.36. Edit CV Languages 3
Back in the main CV dialog, you will now find the list of languages under the current CV (1). There is also a second column with the language label under 'Entries' (2). Select the language from the list (1) you want to add, change or delete entries for.
Figure 2.37. Edit CV Languages 4
By clicking an empty field in the entries (2), you can now enter the required values (3). To add a new entry, enter the values and click Add (4). This will yield a new line in the entries. To add a value to an existing entry, as shown in the screenshot, click Change. Lastly, you can delete entries as before with the Delete button.
If you use a lot of multi-lingual CV's, you can also set the preferred default language to work with. More info about this can be found here: see Using CV's
This functionality is currently out of order, please see the note at Data categories
In ELAN it is possible to associate an entry of a controlled vocabulary with a data
category of the ISO Data Category Repository (DCR). To do so click on
Edit in the ELAN main menu and select Edit
Controlled Vocabularies.... In the window that now appears, click the
Browse... button to the right of the ISO Data
Category
option. In the next window, select a category from the list and
click Apply. Or you can add a category by clicking
Add Categories; first select a profile in the left panel
and then select a data category in the middle panel. Next, click on Apply to put the
selected data category in the list of selected categories. Then click to highlight
it
and click Apply to associate it to the CV entry.
More information about the ISO DCR and how to use it can be found in Data categories.
Instead of creating a Controlled Vocabulary again every time ELAN also supports the
possibility to reuse existing CV's. A first way to achieve this is to use a template
file that contains a CV when creating a new *.eaf
file. (See Creating a new document) Alternatively you can import a CV for already
existing files:
A created or imported controlled vocabulary is part of the current transcription and
is therefore part of the .eaf
file when the transcription is saved.
However, there are situations in which a controlled vocabulary is shared between two
or
more transcriptions and change to that CV should be propagated to all transcription
that
use that CV. For this purpose, ELAN enables you to link to a controlled vocabulary
that
is either on your local or attached hard disk or somewhere on the network or the Internet:
Similar to the Import CV process, if you try to import a CV with the same name as an already existing CV a dialog will pop up asking what to do (see Controlled vocabulary already exists warning).
The entries of an External CV cannot be edited nor their ISO Data Category. Also the order of the entries cannot be change. The possibility to add a color and shortcut key to an entry via the More Options... button is still there.
A CV can be exported by clicking the export .ecv button in the Edit Controlled Vocabularies window. This will open another window, in which you can select the CV's to be exported:
Figure 2.40. Export Controlled Vocabulary
Finally, select a location to save the CV. The file will have the .ecv extension.
Open the Change the Tier type dialog (via Type > Change tier type) and make a choice from the Use Controlled Vocabulary dropdown box:
Figure 2.41. Change tier type
When editing an annotation that belongs to a tier type associated with a CV, a drop down box appears, the suggest panel. Just select the value to be filled in:
Figure 2.42. Controlled vocabulary drop down box
If you have a created a multi-lingual CV, you can set the preferred language to use by going to Options > Language for multilingual content. From there you will see a list of languages that you have used with a CV. Select the one preferred and afterwards, values you enter are in the CV-language chosen. When the default language changes, all Annotations which are associated with a CV entry will be adapted (if the selected language actually exists in the CV, and if the entry is not empty in that language).
Figure 2.43. Select language for multi-lingual CV
More information about entering annotations from a CV and working with the Suggest Panel, see Entering annotations from a Controlled Vocabulary
The CV entries for the child annotation value can be filtered/suggested based on the parent annotation value. To enable this,a preference setting has to be enabled. By default, this preference is unselected. Edit -> Preferences -> Edit Preferences -> CV -> Suggest child annotation CV entries based on the parent annotation value
The child tier should have a tier type associated with a CV and a symbolic association stereotype. The automatically filtered/suggested values depend on the two options:
If this option is selected, the suggestions are based on all individual words of the parent annotation value. The result also is in combination with Suggest Panel options selected in the CV preference.
Figure 2.44. Look for CV entries that contain instead of start with the search string unselected
Figure 2.45. Look for CV entries that contain instead of start with the search string selected
The suggestion also considers Also look in the descriptions of CV entries and Ignore case
If this option is selected, as the name suggests, suggestions are based on the exact sequence of words of the parent annotation value.
Figure 2.46. Suggestions based on the exact sequence of words
The suggestion also considers Also look in the descriptions of CV entries and Ignore case