Skip to main content
Skip table of contents

Vocabulary Widget


The Vocabulary Widget allows you to easily add Data Classification capabilities to your data capture tools by drawing on vocabularies hosted at Research Vocabularies Australia.

The widget has been written in the style of a jQuery plugin, allowing complete control over styling and functionality with just a few lines of JavaScript. The widget also ships with some UI helper modes for:

  • searching for vocabulary terms, with autocompletion;

  • populating a select list (or autocomplete textbox) with items derived from a base IRI (which is either the broadest term of a hierarchy, or a collection);

  • browsing a hierarchical vocabulary as a tree.

It is also possible to use the widget in a more programmatic manner; refer to the Advanced Configuration section below for more details.

How do I use the widget?

Pull in the widget code and styles

The widget requires jQuery; load this, and the plugin itself (and associated CSS styles) in your document’s <head>...</head> segment:

<script src="//"></script>
<script type="text/javascript" src="//"></script>
<link rel="stylesheet" type="text/css" href="//" />

If you are using tree mode (explained below), you may wish to display the widget in a tooltip. If so, add this to pull in support for qTip2 tooltips:

<script type="text/javascript" src="//"></script>
<link rel="stylesheet" type="text/css" href="//" media="screen" />

The code generated for tree mode by the Widget Explorer requires qTip2. When incorporating the results of the Widget Explorer, make sure to include the content from both of the code blocks given above.

Instantiate the widget

Add widget input and initialisation code into your HTML document’s <body>...</body> section. The input/initialisation code selects the vocabulary, input mode, and other parameters. Write the code yourself, based on the instructions in the following sections, or use the Widget Explorer to generate it for you.

You can combine the two approaches: use the Widget Explorer to generate boilerplate input/initialisation code, then customise it to suit your needs.


This section gives one demo of each of the main supported modes.

Search mode

Search a vocabulary for matching terms, provided in an autocomplete-style list:

Narrow and collection modes

Narrow and collection modes can be attached to a select element to provide a drop-down, or a text input box for an autocomplete-style list:

  • Narrow mode supports selection from the concepts that are narrower than a specified concept (to be specific, the concepts that are skos:narrower than a specified skos:Concept).

  • Collection mode supports selection from the members of a collection (to be specific, the skos:members of a specified skos:Collection).

Narrow mode example

This demo uses a <select> element.

Collection mode example

This demo uses an <input> element that supports autocompletion.

Tree mode

Tree mode constructs a clickable vocabulary tree for a given repository. Bind to the treeselect.vocab.ands event to handle user selection.


The plugin accepts a suite of options, detailed below. Please note that some options are required, and don’t have default values (such as repository): you must provide values for such options. Incorrectly configured plugins will result in a JavaScript "alert" box being displayed, describing the nature of the configuration problem.

Options are passed into the plugin as a JavaScript object. For example:

$("#vocabInput").vocab_widget({cache: false});

Be sure to quote strings, and separate multiple options with a comma (,).

Alternatively, options can be set after initialisation using the following form:

$(...).vocab_widget("[option-name]", [option-value]);

This works for all options except mode, which must be specified at initialisation (or omitted for core usage).

Some options are specific to the chosen mode; the tables below are grouped in a way that makes this easy to comprehend. Core usage of the widget exposes all “common” options.

Note: tree mode has no specific configuration other than the widget’s common options.

Common options






The API key you have previously been issued.

Please request an API key by emailing



REQUIRED Vocab widget mode:

  • "tree": provides a hierarchical display based on broader/narrower relations

  • "search": provides an autocomplete widget on an HTML input element

  • "narrow": populates an HTML select element with appropriate data based on narrower relations.

  • "collection": populates an HTML select element with data from a collection.

  • "advanced" or "core": exposes the core widget with no UI helpers.



REQUIRED The SISSVoc repository to query. This can be specified either as:

  • a base repository name, e.g., "anzsrc-for""rifcs"

  • a URL that is the prefix of the SISSVoc endpoints for the vocabulary, e.g., "↵



At most, how many results should be returned? Note: the vocabulary's SISSVoc configuration may override this option by applying a lower (but not a higher) maximum limit. At present, the SISSVoc configuration of most vocabularies specifies a maximum limit of 200 results.



Cache SISSVoc responses?


"ARDC Vocabulary Widget ↵
service error."

Message title to display (via a JavaScript alert() call) when an error is encountered. Set to false to suppress such messages



Location (absolute URL) of the "back end" of the widget (an interface to SISSVoc that generates responses in JSONP).

Search mode options

In this mode, the target element should be <input type="text" ...>.






How many characters are required before a search is executed?



How long to wait (after initial user input) before executing the search? Provide in milliseconds


"No matches found"

Message to display when no matching concepts are found. Set to false to suppress such messages



CSS ‘class’ references for the dropdown list. Separate multiple classes by spaces


["label", "notation", "about"]

Which fields do you want to display? Available fields are defined by the chosen repository.



What data field should be stored upon selection?

Narrow and collection mode options

In these modes, the target element should be <select ...>.






REQUIRED In these modes mode, mode_params defines the vocabulary item upon which to narrow. The value is the IRI of the vocabulary item.


["label", "notation", "about"]

In these modes, if the target element is a <select>, this option must be overridden to be an array of one string (e.g., ["label"]). This selection defines the label for the select list options.



What data field should be stored upon selection? This field is used as the value attribute for the select list options.

Advanced Configuration

Core mode

Invoking the plugin with no mode argument, or with the mode set either to "core" or "advanced" exposes core functionality, without having to use form input (text, select) elements or the like. Instead, you hook into JavaScript Events, building the UI as best fits your needs. A very basic example is shown below: it constructs a list of RIF-CS identifier types.

This form section is using the widget with no helpers; it outputs a list of known RIF-CS identifier types:

Core mode exposes the functionality that underpins the other modes, making it accessible through calls to the vocab_widget() function. Having initialised the widget in core mode, you call vocab_widget() again, passing in one of the following values as the first parameter:

  • "search"

  • "narrow"

  • "collection"

  • "top"

You also pass in a second parameter, which can look like either of the following:

  • a string: search term (for a "search" call) or narrow URI (for a "narrow" call)

  • an object {uri: "...", callee: my_callback_object }:

    • The uri provides the same data as per the description of the string value in the point above.

    • The callee defines the object against which the subsequent JavaScript event will be triggered. Defaults to the containing element (i.e., on which you invoked vocab_widget()).

When you call the vocab_widget() function in this way, a request is made to the widget endpoint (the "back end" of the widget) to fetch data from the vocabulary. When the endpoint replies, an event is generated and the response data passed in as the event's data. For example, if you call vocab_widget("narrow", ...) there will be a narrow.vocab.ands event generated when the endpoint replies.


Events are fired to allow you to hook into the widget workflow and implement your customisations as you see fit. Each of the widget's modes supports configuration of event handlers, but only certain events are fired in each mode. For example, the treeselect.vocab.ands event will only be fired when the widget has been configured in tree mode. For modes other than core mode, there are default event handlers already defined, and you can override them. For core mode, there are no default handlers; you need to define handlers for each back end function you are going to request using the vocab_widget() function as explained in the previous section.

Widget events are in the vocab.ands namespace. This “legacy” name has been retained in order to continue to support existing users.

Event Name




  1. JS Event object

  2. SISSVoc data object

Hook into the plugin’s search core function; data is the search response.


  1. JS Event object

  2. SISSVoc data object

Hook into the plugin’s narrow core function; data is the search response.


  1. JS Event object

  2. SISSVoc data object

Hook into the plugin's collection core function; data is the search response.


  1. JS Event object

  2. SISSVoc data object

Hook into the plugin’s top core function; data is the search response.


  1. JS Event object

(Fired in search mode, and in narrow and collection modes when the target is an <input> element.) Fired when a search result item is clicked. The selected item is the event target. The target will have a vocab data object, containing all the details found in a SISSVoc data object.


  1. JS Event object

(Fired in tree mode.) Fired when a tree item is clicked. The selected item is the event target. The target will have a vocab data object, containing all the details found in a SISSVoc data object.


  1. JS Event object

  2. XMLHttpRequest

This event is fired whenever there is a problem communicating with the plugin’s endpoint (the widget back end).

If the error occurred during an AJAX call, the object will be a bona fide XMLHttpRequest object. Otherwise, a plain object with status and responseText properties will be available.


The SISSVoc data object returned by the above events (and also attached to the searchselect and treeselect event’s vocab data object) is a plain object with the following properties:

  • status: OK if all good, something else (most likely ERROR) if not

  • message: description of the underlying system call by default, or information on status when something went wrong

  • limit: the maximum number of records requested

  • items: an array of SISSVoc vocabulary items:

    • definition: item description

    • label: item label

    • about: item definition / URL

    • broader: parent term (if it exists)

    • narrower: child terms (if they exist, otherwise boolean false)

    • count: frequency of use among ARDC registry objects (experimental; requires endpoint to be set to ""; works best on ANZSRC-FOR, not so well on RIF-CS)

  • count: the number of items returned

JavaScript errors detected

Please note, these errors can depend on your browser setup.

If this problem persists, please contact our support.