Explore, Query & …. Visualize

Posted by

What is the real difference in value between “search of strings” and “search of things”? The ability to quantify, I would say. When text based search is able to provide a SEO manipulated list of best guesses, it can not with any certainty answer questions like “how many”, “how much”. And these are the interesting follow-up questions in any business after the initial first “is there” quick search.
The value of the structured search (where semantic search in linked data belongs) is the ability to quantify and therefor it is actually surprising how little effort there is put in SemTech industry and community to go beyond of list based results and focus on one “Thing” at the time. By this “one thing” I mean the multitude  of efforts to provide richer Linked Data browser “describe” pages, focusing on that one current thing – its properties and relations.

Dat-a-Mazehouse framework provides the integration with Spotfire web player  which allows instant visualization of SPARQL query result sets for commercial users and Spotfire clients. It is not only about static one time display of charts but interactive discovery and enrichment tool. The visualization contains facets and filters which can be applied to further refine and understand the aspects of “how many” questions. The selected filters can be then taken back to and added to the initial SPARQL query and thereby reduce the amount of data relevant to the analysis. Visually marked items on any graph can be grabbed and typed in the triple store…

But this blog is not about the power of Data-a-Mazehouse and Spotfire integration but about Explore&Query and Sgvizler. The introduction provided is to present the context of value of SPARQL + Visual analysis.

Sgvizler is a contribution of Martin G. Skjæveland.

Bob DuCharme covered it in his recent blog entry. As Sgvizler is provided under free MIT license, it is a perfect fit to combine with Explore&Query free edition and turn this tool into E,Q & Visualize.

So when you have built the SPARQL query by browsing through the linked data

I

You can pick an appropriate graph type from the list (plenty to choose from!)

II

… press the Run button and you get an instant insight to your data and answers to your “how many” questions.

III

With geo coordinates available in the data you can even get the instant answers to the questions like “how many where”

IV

Welcome to download your copy at Explore&Query.