A lightweight, backend-free open data portal, powered by Jekyll
HTML CSS JavaScript Ruby
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
Research
_data
_datasets
_includes
_layouts
_organizations
css
favicon
img
scripts
.babelrc
.gitignore
.travis.yml
01_index.html
CONTRIBUTING.md
Gemfile
Gemfile.lock
LICENSE
README.md
Rakefile
_config.yml
_config_localhost.yml
add-dataset.html
add-organization.html
admin.html
data.json
datasets.html
datasets.json
favicon.ico
manifest.json
organizations.html
package.json
resources.html
setup.md
webpack.config.js

README.md

My take on the design of JKAN - a lightweight, backend-free open data portal, powered by Jekyll. Learn more about JKAN in the original .readme here.

Read more about the project idea and history here

And check out the existing design

Get started

To start the site, run gem install bundler in the project directory. Once it is complete, run rake run and when started go to http://127.0.0.1:4000/data-portal/ in your browser.

Rebuilding assets for production builds

If you've edited any of the static assets, you may need to run npm run build to rebuild the asset bundle to get those changes to show up up on the github-pages hosted site. After building the bundle, commit the bundle file(s) (i.e. bundle.js) and push to github to make the changes live.

Core Purpose of Site

Enable folks to learn about, discover, and post links to open data.

User Testing:

Notes on similar platform from Code for America Regional Data Portal Webinar

  1. "It's tough to know where to start as a data user"

    • Overwhelm homepage with newb stuff
    • Trim down categories for data explorer
    • Distinct flows for posting & getting
  2. "It's challenging to explain scientific data to non-scientists"

    • State purpose of collection
    • Explain what the metadata means
    • Suggest usage
  3. "It's difficult to download an entire dataset and make it useful"

    • Give examples of use
    • Encourage documentation & walkthroughs
    • Encourage guest lectures at meetups

SWOT

  • Strengths
    • Streamlined
    • Simple
    • Lightweight
  • Weaknesses
    • Unwelcoming
    • Assumes prior knowledge
    • Doesn't demonstrate value
  • Opportunities
    • Welcome users
    • Demonstrate value of open data
    • Help with projects
  • Threats
    • Licensing Issues w/ Datasets
    • Complexity
    • Scope Creep

Design Challenge:

Make the data portal newbie friendly without disrupting the flow of existing users.

Make the design modular so that other cities can fork & replace our content.

Sitemap

Before

After

User Stories

As a newbie, I can arrive at the homepage, read about what's going on, see some examples, and feel like I know how to get started.

As a visitor, I can go straight to the portal page to accomplish my task.

As a visitor, I can browse datasets by category and contributor.

As a visitor, I can view projects which correspond to a dataset.

As a visitor, I can figure out how to become a contributor.

As a visitor I can view a resource page to learn how to structure & store data, what tools to use to work with it, & how to get a project off the ground.

As a visitor, I can mark that I've used a dataset.

As a visitor, I can leave an evaluation for a dataset.

As a contributor, I can post a dataset.

As a contributor, I can link to a project I made using a dataset.

Inspiration

Staying consistent with Open Austin branding makes a lot of sense.

The Opportunity Project provides examples of projects to demonstrate the value of open data.

The City of New York data portal is built on socrata but has designed around that. They must be doing something right since they are the #1 place in the US for open city data.

Sketches

I decided to build out a homepage and resources page in addition to the portal.

Small

Medium

Large

Wireframes

Small (Individual files here)

Medium + (Individual files here)

Mockups

Small (Individual files here)

Medium + (Individual files here)