resource-align

DEPRECATED (RETIRED) This repository is now retired, achived and no longer maintained - note particularly that security updates are NO LONGER being performed.

Web service to align resources to curriculum standards, drawing on multiple criteria

NOTE: This is experimental and proof-of-concept code

This code builds on https://github.com/nsip/curriculum-align (which it also invokes), and it proposes resources from a repository that are aligned to a given set of learning areas, year levels, and (optionally) curriculum standards, in ranked order.

Binary distributions of the code are available in the build/ directory.

The web service is made available as a library (Align()); the cmd directory contains a sample shell for it, which is used in the binary distribution. In the sample shell, the web service runs on port 1576. The test script test.sh issues representative REST queries against the web service.

The web service takes the following arguments:

GET http://localhost:1576/align?yr=X1,X2,X3&area=Y1,Y2,Y3&item=Z1,Z2,Z3....

where yr is the year level, area is the learning area, and item is the curriculum standard to be aligned. All of these can have multiple comma-delimited values. All parameters are optional; if no parameters are given, alignment is attempted against all curriculum items configured in the https://github.com/nsip/curriculum-align instance.

The repository that this implementation works from is a set of JSON files in the repository folder of the executable; in this distribution, sample files are in cmd/repository. The JSON structure required is an array of JSON objects, with the following structure:

• URL: URL of resource
• Content: Textual content of resource, to be used for alignment to curriculum via https://github.com/nsip/curriculum-align. (NOTE: the value of Content needs to be kept to a single line.)
• Paradata: JSON object, mapping curriculum items codes to the frequency with which the resource has been used to teach against that curriculum item in the repository
• Manual-Alignment: List of nominated curriculum item codes that the resource has been aligned to in metadata, whether by the resource author, a local expert, or a third party
• Learning-Area: List of learning areas that the resource relates to; is expected to be the same set of values as used in the related curriculum
• Year: List of year levels that the resource relates to; is expected to be the same set of values as used in the related curriculum

For example:

  {
    "URL":"http://www.skwirk.com/p-u_s-4_u-198/planet-earth/nsw/science",
    "Content":"Chapter 1: The layers of the Earth Chapter 2: The tectonic plates Chapter 3: Volcanoes: birth, life and death Chapter 4: Earthquakes: cause, reaction and measuring Topic 2 : Rocks and minerals Chapter 1: Minerals, crystals and ores Chapter 2: Igneous rocks Chapter 3: Sedimentary rocks Chapter 4: Metamorphic rocks Topic 3 : The atmosphere Chapter 1: The layers of air Chapter 2: Weather and climate Chapter 3: The greenhouse effect Chapter 4: The ozone layer Topic 4 : The hydrosphere Chapter 1: The water cycle Chapter 2: Tides Topic 5 : Shaping the Earth Chapter 1: Weathering: chemical and physical Chapter 2: Erosion: people and nature",
    "Paradata":{
      "ACSSU153":4,
      "ACSIS148":1,
      "SC4_12ES":2,
      "SC4-3VA":1
    },
    "Manual-Alignment":[
      "ACSSU153",
      "SC4_12ES"
    ],
    "Learning-Area": ["Science"],
    "Year": ["7", "8"]
  },

The response is a JSON list of structs, one for each resource matched, with the following fields:

• URL: The URL of the resource
• Content: The abstract of the resource
• Statement: The identifier of the curriculum item for which this resource is the best match. The curriculum items matched against are filtered by the parameters given in the web service call.
• Expert: The score for the best match based on expert advice (the Manual-Alignment field of the repository entry).
• Usage: The score for the best match based on usage (the Paradata field of the repository entry).
• TextBased: The score for the best match based on keyword alignment (calling curriculum-align with the Content field of the repository entry).
• WeightedTotal: the overall score for the best match; currently the sum of the Expert, Usage, and TextBased scores.

Each of the Expert, Usage, and TextBased scores are normalised: for each item, they are normalised to range from 0 to 1. The response is ranked in order of WeightedTotal.

As with curriculum-align, all text tokenised in the gem from the resource repository is also indexed and available for retrieval:

GET http://localhost:1576/index?search=word