Content Details

When parsing content, Generator will extract as much useful information as possible. This includes text, but also other details such as authoring tool version or content length. As we improve our parsers, we will add more information to the content_details property of the content object.

Currently, the content_details object looks like this:

{
  "content_type": "STORYLINE",
  "existing_metadata": {
    "authoring_tool_version": "1.0.0",
    "slides": 7
  },
  "interactions": [
    {
      "question": "What is 6x6?",
      "choices": [
        {
          "text": "36",
          "is_correct": true,
          "category": null
        },
        {
          "text": "66",
          "is_correct": false,
          "category": null
        },
        {
          "text": "12",
          "is_correct": false,
          "category": null
        },
        {
          "text": "1",
          "is_correct": false,
          "category": null
        }
      ],
      "feedback": [
        "That is incorrect. 6x6 is equivalent which is 36",
        "That is correct. 6x6 is equivalent which is 36"
      ],
      "type": "StorylineInteractionType.MULTIPLE_CHOICE",
      "range": null
    }
  ]
}

Existing Metadata

The existing_metadata property defines metadata properties that are extracted directly from the learning content. The schema of property is dependent on the type of content. Details about the supported schemas and their properties are listed below.

Each metadata object has a property intended to signify “length”, which indicates the total number of ’elements’ that will be visible to learners taking the course. Generator makes an effort to align the term used to indicate “length” with the terminology used by each authoring tool (such as slides or pages). Additionally, Generator will consider element visibility when determining the content “length”, as not all elements defined in the course structure will ultimately be visible to learners.

For example, a course may present a quiz with one question per slide. The course may define 10 questions in its question bank, but will only present 5 questions to a learner. For this course, the “length” of the quiz section would be “5”.

When applicable, Generator will also expose the version of the authoring tool used to publish the content. This version string is defined in the authoring_tool_version property. A value of null indicates that the tool doesn’t support or expose versioning (as is the case with Articulate Rise 360 and gomo). If Generator fails to derive the tool version with content from a tool that typically defines a version value, Generator will use the value "undetected".

PDF Metadata

Url Metadata

VTT Metadata

Storyline Metadata

Rise Metadata

DominKnow Metadata

IntellumEvolve Metadata

Gomo Metadata

Captivate Metadata

Interactions

Among other content data, Generator will attempt to extract information about interactions present in the content. Because interactions can vary greatly within and between authoring tools, Generator provides a flexible abstraction that can apply to many different types of questions. The format for interactions metadata is as follows:

• type: The type of interaction, such as “MULTIPLE_CHOICE” or “MATCHING”. Generator will use the terms defined by the authoring tool to describe an interaction.
• question: Usually the text of the question. In some courses, it will simply be a reference or title to the interaction.
• choices: The potential options a learner could choose. Here are the properties for choice objects.
  • text: Text representing the choice (some choices might just be images).
  • is_correct: Whether a choice is considered correct. Note that Generator makes no judgement about whether a choice is necessary or sufficient for an interaction to be considered correct as a whole.
  • category: This is a string to be used by interaction types which involve matching or grouping, rather than simple selection. The string represents the correct group for choices to be assosiated with.
• range: Used for numeric questions or sliding scales. It has two properties.
  • min: The minimum value possible.
  • max: The maximum value possible.
• feedback An array of strings which represent potential feedback text depending on the learner’s answer to the question. Most likely a “correct” and “incorrect” version, but more involved feedback is possible.

Currently, Generator only supports interactions for a subset of authoring tools, but Rustici is actively working on support for the remaining authoring tools. Below is documentation of the interaction types from each authoring tool:

Storyline Interactions

Rise Interactions

IntellumEvolve Interactions