Techquila Blog - Latest Comments

Re: Charityware

Wendy Dessler — Tue, 21 Jun 2016 02:34:23 -0000

Thanks for sharing!Crowdfunding has now become an effective and reliable way of raising funds that a lot of people who are in need are benefiting from it. You can also check http://www.plumfund.com/ and learn more about the benefits of crowdfunding.

Re: Sample Topic Maps

kal — Fri, 17 Feb 2012 11:35:23 -0000

Hi Marcus,

Thanks for getting in touch! You are right, Techquila is not active any more as a business. My time is spent working in my new company NetworkedPlanet. Our Web3 product is a full topic maps engine with a nice schema editor interface and a user-friendly, ontology-driven topic map editor. It can also be used to publish your data as RDF Linked Data with support for SPARQL queries, so it is really a full system for creating and publishing linked data.

I'm also working on BrightstarDB, which is a NoSQL database for .NET that just happens to be built on an RDF triple store. It doesn't have all the nice GUI of Web3, but it has a lot of neat features for developers so you might want to check it out.

Both products have a free download (Web3 has a limited trial period; BrightstarDB is free to use for internal development purposes and open-source projects), so if you want to check either of them out I encourage you to do so!

Re: Sample Topic Maps

Marcus Vinícius Chevitarese Al — Thu, 16 Feb 2012 14:01:09 -0000

Hi, Kal Ahmed.
I'm a system analyst at Brazilian House of Representatives and I'm very interested about using 'topic maps'. We've a lot of information that we need to organize and to show to the citizen in a readable way, besides this information also needs to be machine-readable.
Seems to me that the website 'techquila.com' is not active. Is it true? Are you working in another site? May you suggest another websites like 'techquila.com'?
Thanks in advance.

Re: Measuring The Semantic Web – Measurement Units

Robert Barta — Mon, 21 Dec 2009 17:52:41 -0000

> Without a declarative approach to units you are forced to hardcode ....

Sure, but that has already been done in code. So my question is: "What is the benefit of having a declarative description of the relationship between units?"

So is there some kind of reasoning?

And even if so: I would rather want to see "units" as part of any knowledge model.

Re: Measuring The Semantic Web – Measurement Units

Kal — Mon, 14 Dec 2009 11:36:36 -0000

@Robert: Without a declarative approach to units you are forced to hardcode all possible conversion factors and your application has no discovery mechanism for reasoning about units that have not been hardcoded into the application. I don't think this is a scalable/sustainable solution for LOD. UCUM codifies in prose the sort of things that I think should be codified in a machine-processable manner - unit definitions and conversion factors.

Re: Measuring The Semantic Web – Measurement Units

Robert Barta — Sun, 13 Dec 2009 09:39:50 -0000

Since units are such an intrinsic part of physics, I actually do not think it necessary to have a fully declarative description of units and their meaning.

Personally, I have been quite happy with UCUM:

http://aurora.regenstrief.o...

For the inclusion into Topic Maps, I have simply "evolved" the model to include units into it, and also the notation, and of course the query language.

A "native" solution. It was fast and easy to do.

Re: Measuring Out The Semantic Web – Definitions

Robert Barta — Thu, 10 Dec 2009 18:17:52 -0000

I can see now how the datacube concept has influenced HyTime. The formal definition is pretty similar.

Re: Measuring Out The Semantic Web

Kal — Thu, 10 Dec 2009 15:57:51 -0000

@Lars Marius:

The next post will look at what I think HyTime has got. Then hopefully the one after that will look at how we could make use of what HyTime has got in a Topic Map. I actually think that scope for assignment might be the easiest way to make this work, but doing something with decorating the occurrence type in the ontology to define a default measurement unit for a specific occurrence type might be an interesting way forwards.

@Bob:

QUDT looks like exactly the kind of thing I am thinking about. Probably the only place I would have a quibble (and it is a minor one) is the use of decimal values to express the conversion factors rather than expressing conversion as a ratio of whole numbers. Maybe I don't have to carry on wading through HyTime :)

@Patrick and Steve

The charter for the OASIS TC looks very promising. Of course, it says nothing about Topic Maps, so perhaps the way forward for this blog is to think about what Lars Marius calls "attachment" (good term) of units to measurements expressed in a topic map. However, I've got far enough down the road of reading the right bits of HyTime that I might as well also write up my own findings as well...stay tuned :-)

Plenty to mull over...

Re: Measuring Out The Semantic Web

Steve Ray — Thu, 10 Dec 2009 15:46:40 -0000

You might be interested to know that the ontology community is just now organizing itself to synthesize a "Units of Measure" ontology as an OASIS standard, using the various existing measurement ontologies. See http://ontolog.cim3.net/cgi.... Having said that, we ought to discuss Hytime - I'll bring your blog article to the group's attention.

Cheers,

- Steve

Re: Measuring Out The Semantic Web

Patrick Durusau — Thu, 10 Dec 2009 01:30:08 -0000

Kal,

Excellent! Just excellent!

Err, there is a charter pending for a TC in OASIS that is going to be working on measurement ontologies (although not the sort of mapping you are talking about). Still, it may be of interest: OASIS Quantities and Units of Measure Ontology Standard (QUOMOS) Technical Committee, http://lists.oasis-open.org....

Where to start? That's a hard question, particularly given my long time co-editorship with Newcomb. Not sure I read it the same way I would have without that. I am willing to look at it again, this month in fact, to see if i can suggest some likely chunks that avoid some of the added complexity. If nothing else it would be more interesting that some of the other reading that occupies a good bit of my time.

This is both interesting and important, not too often that is seen in any area of endeavor.

Hope you are having a great day!

Patrick

Re: Measuring Out The Semantic Web

Bob DuCharme — Thu, 10 Dec 2009 00:02:07 -0000

Hi Kal -

I don't know if you heard that I've joined TopQuadrant last August, but around that time some co-workers convinced NASA to let them publish the measurement units ontology that they developed with NASA. See http://composing-the-semant.... I'm sure they'd be very happy to get some feedback on its use.

Bob

Re: Measuring Out The Semantic Web

Lars Marius Garshol — Wed, 09 Dec 2009 20:10:07 -0000

First of all, I think you are right that this is a major issue, and particularly in a linked data context. Some people try to handle this by putting the unit into the occurrence type name or in documentation on the occurrence type (or property in RDF), but it's difficult to argue that this is a good solution.

I'm not convinced that HyTime has much to offer here, though. It may have some text about N-dimensional spaces and so on, but does it have an actual syntax and actual algorithms for handling this stuff? Does it have any way of detecting that your and my independent definitions of ISO meters are the same? If so, what does all of this look like? (A blog posting answering this would be nice. :)

However, let's say that HyTime does let you express this, and that it does so in a way that's better than simply referencing known units like celsius, kelvin, fahrenheit, reaumur, etc. Then what? If your nice HyTime SGML document sits over there, and your nice topic map is here, what good is that? What I'm getting at is that there must be some way to connect the actual number in a data set with the definition of some unit. That may well be harder than coming up with definitions for the main units in use.

There are some ways one could do this:
* Topics for units, then scope for attachment. (Meh.)
* Topics for units, then reification for attachment. (Double-meh.)
* Topics for units, then standardized association from occurrence type for attachment. (Well...)
* Use subtypes of datatypes. That is, unit:length-in-meters subtypes xsd:float.

The last one in many ways seems like the best, but it does require you to have some machinery for expressing subtyping of datatypes. Plus you probably need to make a topic for your data type and associate with your unit of choice. Given all that it might be workable.

Re: New NPCL Editor Released

Kal — Tue, 24 Mar 2009 08:00:52 -0000

Hi Diego,

Thanks for your comments!

I never had enough time with TMShare to get something releasable. It was done so long ago and using old technologies such as JXTA that I don't think that the source would be very interesting, probably the most interesting part of TMShare was the topic map fragment exchange protocol described here in the paper I wrote.

Re: New NPCL Editor Released

Diego Vera — Sun, 22 Mar 2009 20:41:27 -0000

Hello there,

I am reading a log your blog and I would like to say that it is a great blog, congrats. I am a researcher in the area of TM and I would like to test TM4J and TMShare, however I could not find TMShare. Do yout know where could I download it? Is it still under development?

Regards,

-Diego

Re: rdfQuery + OpenCalais + Cloud Storage = Personal Knowledge Base ?

Kal — Wed, 11 Mar 2009 18:25:06 -0000

Hi Tom,

Thanks for the comments. You are right on the money about facts and events - they are definitely at least as interesting as entities, maybe more so because they form a nexus for entities and this mechanism could be used to capture opinions/assertions about the interactions between entities and events.

As for a demo...well there is a BarCamp coming up in Oxford fairly soon and I'm toying with the idea of pitching for some help with this there. But maybe it would be worth my while getting some basic crufty demo together before that and then letting the Javascript/design gurus make it work better/look prettier. If only there were 24 more hours per day...

Re: rdfQuery + OpenCalais + Cloud Storage = Personal Knowledge Base ?

Tom Tague — Wed, 11 Mar 2009 17:52:05 -0000

Kal:

Tom Tague from Calais here.

Some very interesting ideas in here. A couple of additional points to think about:

First, yes Calais does provide consistent identifiers. But one thing to keep in mind is that those identifiers are only disambiguated for a subset of the entity types we recognize (Company, geography, etc). So "John Doe" "John Doe" with any degree of certainty across sites. On the other hand "IBM" = "Taligent" = "International Business Machines" across all sites.

Second, don't forget about facts and events! While entities are great - being able to create community / organize information around events (e.g. XYZ Corporation announces Bankruptcy, a Tsunami occurred in Tibet) could be equally interesting.

Can we take it beyond a personal knowledge store to a shared store with personal spaces? The collaboration opportunities would be pretty great.

When can we expect a demo?

Re: Long time no blog

Kal — Sat, 28 Feb 2009 01:17:37 -0000

Heh...Anyone else out there still have this RSS feed bookmarked. It would be interesting to hear.

Re: Long time no blog

stefan lischke — Fri, 27 Feb 2009 23:53:52 -0000

hi kal thats strange, i must have subscribed to your blog over
4 years ago so i got your update today. time is running so fast
cya
stefan

Re: Pepys-Map: July and edits

Lars Marius Garshol — Tue, 03 Aug 2004 09:58:56 -0000

Thanks for the .zip! I've been waiting for that, since I wanted a single-file download that's got the whole thing in it. Now I'm off to Omnigate. :)

Re: Guest Editor!

Alex — Mon, 26 Jul 2004 05:39:48 -0000

Welcome Stuart to the most exciting happening in the TM scene for a loooon time. :)

Re: Pepys-Map: 7th July 1661

Lars Marius Garshol — Fri, 16 Jul 2004 15:22:43 -0000

Actually, I think you want *less* specific role types, rather than more specific ones. My experience has been that associations and even more so roles should have as general types as possible. So I would make it something like "officiator", which would work nicely for other church seremonies as well, but wouldn't fail to work in, say, a mosque. (Not that Pepys ever entered one, but still.)

Re: Pepys-Map: 2nd July 1661

Steve Pepper — Thu, 15 Jul 2004 16:49:41 -0000

I don't know much about The Siege of Rhodes. I don't think it was the first opera to be staged in Britain (England?), but I think it is regarded as the first English "opera". Whether it was really an opera is a matter of definition, I think. It seems to be more of a play with musical interludes and a bit of recitative. Perhaps more like the German Singspiel. I think there is also an interesting angle relating to Puritanism (all of Pepys diary is taking place during the English Revolution, remember): I believe "plays" had been banned for being amoral, and calling something an "opera" was one way of circumventing the ban.

Re: Pepys-Map: 2nd July 1661

Kal — Thu, 15 Jul 2004 03:18:19 -0000

Thanks Steve! I've added Sir William to the people topic map and added a composed-by association with "The Seige of Rhodes".
I read somewhere that this was the first opera staged in Britain...care to comment on that ?
Updated topic maps from the same links above (I've fixed the link to the diary entry topic map).

Re: Pepys-Map: 5th July 1661

Kal — Thu, 15 Jul 2004 03:00:09 -0000

Pepys certainly seems to get through a lot of venison pasties.
Perhaps there should also be a "being-merry" event ...that seems to happen a lot too... :-)

Re: Pepys-Map: 5th July 1661

Steve Pepper — Thu, 15 Jul 2004 02:05:39 -0000

Hey! Getting lazy? What about an eating-event for the venison pasty :-)