Lightweight data browsing with Miga

From semantic-mediawiki.org
Jump to: navigation, search
SMWCon Fall 2013Lightweight data browsing with Miga
SMWCon Fall 2013
Lightweight data browsing with Miga
Talk details
Description: How to run and use the Miga Data Viewer, and how it fits into the Semantic MediaWiki ecosystem.
Speaker(s): Yaron Koren
Slides: File:SMWCon Fall 2013 MigaDV presentation.pdf
Type: Talk, Demo
Audience: Everyone
Event start: 2013/10/30 05:00:00 PM
Event finish: 2013/10/30 05:20:00 PM
Length: 20 minutes
Video: click here
Keywords: Mobile, Semantic Drilldown
Give feedback

The Miga Data Viewer, or Miga for short, is a lightweight JavaScript framework for displaying and browsing structured data that's stored in CSV files, that I released in July 2013. It does not directly use or interact with Semantic MediaWiki, but it has a number of important connections to SMW:

  • It was inspired by the generic approach to data structures pioneered (or at least popularized) by SMW.
  • Its approach to data storage borrows heavily from SMW, including storing data as triples, and a similar set of data types (Text, Number, etc.).
  • Its browsing interface borrows heavily from the Semantic Drilldown extension.
  • It includes some "importer" scripts for creating CSV files from various data sources, including one for SMW (the other two are for MediaWiki and Wikidata).
  • It can function as a mobile-friendly interface (with offline capability) for viewing the data from an SMW wiki.

Additionally, the "smart" approach to data browsing that Miga takes suggests possibilities for improvement of the SMW ecosystem. It has already led to one major change: in Semantic Drilldown, as of version 1.3, filter values are generated automatically, based on the data layout; they are no longer defined manually. There are other potential improvements that could be made:

  • Perhaps there should be maps automatically generated for pages with coordinate properties.
  • Perhaps there should be calendars/timelines automatically generated for pages with date properties.
  • Perhaps every category should have a table showing all its data automatically generated.
  • Perhaps every property should automatically display a distribution graph of its values.

It's possible that the Semantic Drilldown extension should be the interface for all these automatic visualizations - essentially replicating the Miga interface to the greatest extent possible. That's a matter of opinion, and potentially there will be an opportunity to discuss this during the talk.