How to build a data first media platform with CKAN

20
Feb

Challenge-bar

Our own Beta platform for CKAN Galvanize is currently being used by the ACT Government in Australia.  Link Digital, in response to the Digital Canberra Challenge, are building a Data First Media Platform (DFMP). This is based CKAN at it’s core and integrated with a number of AWS services and the Drupal web content management system.

What is the Digital Canberra Challenge?
It is an ACT Government initiative to engage Canberra’s brightest minds to help build and support a dynamic, healthy and prosperous Digital City for the 21st century.

Link’s Challenge
To create a whole of government image and video library which can be used not only by ACT Gov employees but the private sector and the public. The image & video library would ensure consistency across the service & reduce costs for a number of items including data storage, ability to browse and share, licensing and talent arrangements, copy right and acknowledgment, archiving and staff resources.

Link’s (short) Pitch:

Our understanding is that an open data platform creates an economic good with the dual benefit of unlimited multi-tenant users and n+1 reuse of media items that might otherwise be created for a single primary use. Providing universal access to this media content creates the opportunity for private-public goods to be created within secondary markets to grow the economic value of the platform. That may result in higher value public or social goods, but it can also result in innovative commercial goods that benefit multiple sectors within the ACT economy.

By having core functionality centralised around low cost media storage, attribution, governance and access services, the ACT Government investment targets only the essential parts of service delivery. An open data approach provides for a core platform around which innovative people and organisations can attach additional services to meet specific needs over the long term.

The components being established to demonstrate uses cases for the broader concept of a public image and video library are as follows.

DFMP Component

The DFMP is the CKAN based component of the site. CKAN will handle access control and metadata storage of the media assets.

CKAN includes a web based uploader of CSV files and a way to find datasets based on browsing and searching.

Some extensions may need to be built to meet the management use cases of a DFMP. Specifically, to upload and browse data resources that are collections of references to images and video resources.

Public Video and Image Library USER Interface Component

A Drupal site will sit alongside the DFMP and will be available to support some of the use cases and provide a richer user experience ontop of the CKAN programatic interface.

Web Based Contribution

This will be a CBR branded page with a workflow that is more suitable for a government user of the Public Video and image Library. This will be the primary interface for contributors.

Features include:

  1. Request and maintain access priviliges to contribute media assets.
  2. Collect mandatory metadata easily through a rich and intuitive interface.
  3. Collect automatic metadata like Location from EXIF embedded within photos
  4. Reliably upload large broadcast/print quality assets to contribute to the Library.

Search and BROWSE

The search and browse component will be a CBR branded page with a workflow allows the general public to access the Public Video and image Library. It demonstrates the way in which the DFMP is inter-operable with any kind of search process (such as Trove).

Features include:

  1. Browse the different collections that make up the library
  2. Faceted search/browsing to allow you to narrow down your current view based on groups within current results (locations, authors etc.)
  3. Search using text for matches in text fields

Storage Infrastructure Component

Will be responsible for transcoding large files and managing archiving of lesser used resources.

This component will be achieved using Amazon S3 with a storage retention policy and Amazon Lambda for triggering transcoding.

Automation Components

Alongside the manual contributions of users via the web based interface, there is also the ability to collate media assets from social media platforms automatically. This also demonstrates the ability to bulk upload assets from external systems.

Flickr automation

The Flickr automation component will provide a way for a super administrator to designate what Flickr pools will be selected for harvesting and inclusion within the DFMP. It will also place records into the appropriate organisation>sub organisation>dataset>resource>recorod for each harvested item.

Twitter automation

The Twitter automation component will provide a way for a super administrator to designate what Twitter tweets will be selected for harvesting and inclusion within the DFMP. It will also place records into the appropriate organisation>sub organisation>dataset>resource>recorod for each harvested item. One or more of the following three filters will be demonstrated. Harvest by hashtag (specifically #CBR), harvest by twitter handle (specifically @abcnewscanberra), harvest by location.

 

Leave A Reply

Your email address will not be published. Required fields are marked *