sds's Introduction

SDS (Simple Dataset Syndication)

SDS is a simple protocol/standard/specification for describing datasets. It should be pubished alongside datasets so taht consumers of the datasets (both human and machines) can understand the metadata of the dataset in a standard way.

NOT EVEN A DRAFT

This is just an idea at the moment. We want to create a community around this idea and help make it a real standard.

Get Involved

Email: ??

What are Datasets?

Datasets are any structured, non-API sources of data. For example:

CSV file
JSON feed
RSS feed
XML document
Shapefile
??

This needs more definition. It is difficult to determine where an API begins and a data feed ends. This could be helpful: http://en.wikipedia.org/wiki/Data_set

Example

The following is an (brainstorming) example of an SDS feed in JSON for a CSV file.

  {
    "sds_version": "0.1",
    "set_name": "Our Example Set",
    "set_description": "Our Example Set is pretty great.",
    "datasets": [
      {
        "type": "csv",
        "url": "http://example.com/datasets/data.csv",
        "name": "CSV File",
        "description": "CSV is still useful and simple.",
        "author": "Author"
        "origin": {
          "author": "Example Original Author",
          "organization": "Example Org",
          "url": "http://example.gov/dept/data-service/"
        },
        "fields": {
          "field_1": "Description of field 1",
          "field_2": "Description of field 2",
          "field_3": "Description of field 3",
          "field_4": "Description of field 4"
        },
        "type_info": {
          "version": "CVS version X",
          "delimiter": ",",
          "line_separator": "\r\n"
        }
      }
    ]
  }

Recommend Projects