Hi Simon,
this is a very interesting topic. If you wondered why there is no
response to your question yet, it's probably the overwhelming scale of
options and tools in this space. I haven't looked into the current BI
landscape for a while. From past experience, all these tools can take
tabular data (e.g. CSV) as input. I would also expect that they have at
least plugins that can digest JSON. Any of these can be produced with
TopBraid, e.g. using the SPARQL endpoint, SPIN templates or (recently)
GraphQL servlets.
My *hope* is at this stage that BI tools include better GraphQL
connectors in the future. The advantage of GraphQL is its introspection
support, allowing a tool to query which fields are available. So in
order to build a chart, a user interface could point at a GraphQL
endpoint, and the BI tool could ask for available fields, then the user
could pick which fields to use for the various dimensions of the data
(e.g. X and Y axis, labels). If that existed, then TopBraid's GraphQL
service would be just one of suitable data sources, and users could
expose the data as shape definitions, see
https://www.topquadrant.com/technology/graphql/
Do you see this as a viable approach for the future?
A more integrated alternative is also possible - having chart/graph
builders available within TopBraid itself. We do include various chart
components, but they'd need SWP-based programming to be embedded into
the UI. While we have a much steeper climb ahead to rebuild all the
infrastructure and features that professional BI tools already offer, we
certainly regard such visualizations as important and would be happy to
learn about your requirements, if you want to discuss further.
Holger
On 22/03/2019 2:11 pm, Simon Opper wrote:
Hi folks
Holger, you had a post on your blog back in 2010 on driving charts
(google based) with SPIN here
<https://composing-the-semantic-web.blogspot.com/search?q=chart>.
I'm wondering where things for this use case of semantically aware
numerical visuals are at now given current SHACL, javascript, D3,
JSON-LD and graphql work.
I've just finished the development of substantial Microsoft Power BI
project that successfully built yet another isolated and messed with
data set and a large body of fragile and non-reusable numerical
queries for visuals.... all the while since I work in Topbraid for
other things, knowing full well the potential of semantics for many
aspects that Power BI nicely ignores.
So.. hoping.. dreaming of not repeating the process I'm on the hunt
for a way to deliver semantically aware visual and UX. I note that for
form based property input aspects, I'm working my way into the shacl
constraints / SWP approach. But I'm not seeing the way forward here
to charting/graphing/numerical types of widgets ?
The queries are simple. e.g. counts, sums, time period patterns. With
the exception of some more difficult, but likely still possible DAX
functions, sparql can replicate the calculations.
It's getting a set of user controls, filters and visual types
interacting in a UX environment that is the gap.
Maybe the paradigm has now changed a little with graphql acting as the
more straightforward client side environment, doing all the data work
with the rdf tool set behind the scenes, using all the libraries that
come with things like appllo making visuals easier. Yet this still
leaves the codebase for visuals off in their disconnected, and
not-reusble world. How would the sparql query and the constraints it
comes with affect the UX ?? Can a sparql query such as a calculation
be encapsulated through the graphql or JSON-LD API ?
Any ideas would be great to hear !
Thanks
--
You received this message because you are subscribed to the Google
Groups "TopBraid Suite Users" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to [email protected]
<mailto:[email protected]>.
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the Google Groups "TopBraid
Suite Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
