Dan Keldsen Interviewing Jeff Dirks of SchemaLogic about Taxo 1.0, 2.0, and 3.0, and what SchemaLogic is calling Business Semantics Management
Full transcription if available by following the permalink to the full entry.
Complete Transcription follows:
Dan Keldsen: This is Dan Keldsen, analyst and consultant with the Delphi Group in Boston, Massachusetts, which is www.DelphiGroup.com.
I have on the line with me today Jeff Dirks who is president and CEO of SchemaLogic in Kirkland, Washington, which is www.SchemaLogic.com.
Hello, Jeff. Thanks for joining me for a quick interview on the past, current and future state of taxonomies, what SchemaLogic is calling Business Semantic Management and the larger umbrella that we at Delphi are calling Information Intelligence.
Jeff Dirks: Great, Dan. Thank you for much and I am very happy to be here today.
Dan: We have both been involved in this general world of taxonomy, schemas, etc., for several years now. About four or five years formerly myself as an instructor and consultant, although a good decade or so more informally since I'm an informal kind of guy.
Would you want to give a little bit of information about yourself and SchemaLogic?
Jeff: Yes. SchemaLogic has been around since 2001. Of all of the founding tenets of the company are around the world that taxonomy plays in helping large organizations manage information.
I think what we have found very gratifying in particular over the course of the last year has been almost a new-found embrace with respect to the importance of taxonomy and helping companies solve some fairly significant problems around information management that are really at the fore of issues that companies care about today in terms of competitive advantage and harnessing information as an asset.
In terms of my background, I am a long-time enterprise software person and executive and think we are on the cusp of a brave new world with respect to how technology and taxonomy helps companies with solving what we think are some pretty significant problems that once solved will be great opportunities to grow business and increase competitive advantage.
Dan: Yes, and I think it's interesting that you fired up the company in 2001, which was obviously not the happiest year. It's been a tough couple of years since then and I think it's interesting that there is an overlap between a focus on taxonomy and ontology and information architecture and all this stuff at the same time as having a tough market.
It makes a whole lot of sense for people to be looking at this.
Jeff: Yes, correct, although you could argue that the year 2001 was in the heart of what might be described as a nuclear winter. The founding principles of the company and the founders themselves had a vision in terms of the role that taxonomy played in, as I said, solving some important problems that in the last year and year and a half have come to the fore as a supporting piece of helping companies get on top of this tectonic shift in information management and to some extent driven by the explosion of data and information that every company has within the four walls of the firewall. So we think it's an exciting time.
It's been an interesting journey. You know, we think we are at the crossroads of some significant change.
Dan: So before we get too far into this, just to set expectation for the listeners and eventually readers, since we have this transcribed, there may be some people here, we are making some assumptions that people actually know what taxonomy and ontology, metadata, and semantics are and that is sort of the world that we are calling Taxo 1.0 which we will get to more later in the discussion.
The main idea of this is we are saying that taxonomy and all of these things are key means of allowing organizations to embrace really organizing information and connecting information or content, whatever term floats your boat, unstructured systems in general to structured or data systems as this intelligent sort of database world.
Those that are find themselves in the position of being able to do a lot more with their information intelligently and practically, which is what we are calling information intelligence.
Jeff: Yes, Dan, that's correct. I think that is certainly from our perspective talking to both our customers, prospects and other business partners in the broader information management space is consistent with our view as well.
Dan: With that in mind we are not going to go into a lot of detail defining taxonomies and ontologies. We are just going to assume that people have an idea on this and if they don't, both we at Delphi Group have free white papers on these topics which our listeners and readers could grab from our site as well as training that we do in consulting services.
I'm sure that SchemaLogic has a number of white papers on their site as well and would be happy to talk about this in more detail. So the main topic for this interview is kind of the short story behind the evolution of taxonomies in a modern sense.
Taxonomies have been around for quite awhile, and other organizing methods. Where we are today and how that gets us into the near future. So the short version of the time I will be discussing is what we are calling Taxo 1.0, which is a manual, heavy lifting static sort of world.
Taxo 2.0, which moves into auto categorization, an ever-evolving model of collaboration, participation within a framework so that people and systems can work more effectively together.
Then the near future to the future, Taxo 3.0, which is business semantics management, which is what SchemaLogic is starting to discuss now, and using that as an on-demand service that allows smooth connections between information and data systems and people, all of those being very important.
Before we dive into the longer story behind that, Jeff, if you could provide a view of where you feel that Taxo 1.0 came from and we can quickly walk through Taxo 2.0 and get to the visionary statement of Taxo 3.0 and how this intersects with what you are calling business semantics management.
Jeff: Great, thanks Dan. I would be happy to. From SchemaLogic's standpoint we believe that business semantics management is white space or figuratively white space that global 5000 companies have overlooked in their information management strategies that when addressed we think help enable new business models or improve existing business processes.
We really see collaboration and participation at the heart of what we are calling business semantics management and frankly a key differentiator in the SchemaLogic solution because it allows for us to control the process of how a group, an individual, a community or even an enterprise strives toward the common version of definitions, knowledge, know-how and corporate memory which really in many ways is the codification of competitive advantage for a company.
Once this collaborative mechanism is in place a company can now account for how to manage this constant evolution of the semantics of the enterprise while providing the framework via SchemaLogic that users, communities and systems can subscribe to as a service.
This is the story of Boeing, for example, where the notion of having a central set of definitions that describe their products, goods and services can be centrally managed by their library services organization and then subscribed to by other organizations such as Boeing Images or other organizations that leverage this single source of truth that is being hosted as a service.
That taxonomy is at the core of this is not surprising as it really is as I mentioned the codification of knowledge and know-how over some domain. The problem that we have seen, and certainly having been around since 2001, is that with taxonomy, or Taxo 1.0 as we are calling it, it is very much a static, siloed and brittle view in many ways of the past that is being expressed in taxonomy.
We really view this as being under the control by the tyranny of the few. It's very hard to change and evolve over time. There is a perception of it being very difficult to manage.
There is a perception that there is significant amount of investment that has to be put into building taxonomies that frankly we have seen as a barrier to entry that we are trying to break down with respect to moving into a more dynamic, open and collaborative means to reach common business semantics that can be made actionable and that may vary on a company by company basis.
So from our standpoint we think that this brittleness around Taxo 1.0, if you look at the evolution of taxonomy in technology that this brittleness and kind of siloed effect really open the door for enterprise search and companies looking to enterprise search and auto categorization as a means to really supplant the need for taxonomy.
Taxonomy was so hard and had so much overhead, what if we could look to someone to provide us some sort of natural language processing algorithms that could automatically categorize our information and build us a set of indices that would replace the need for taxonomy or taxonomists.
Because Taxo 1.0 had this perception of slowness, being brittle, being hard and costly to get at, we figured let the search folks capture the hearts and minds of the market that auto categorization really was the silver bullet to replace Taxo 1.0.
We think that notion has begun to diminish and we are not seeing a new view of taxonomy in a broader sense.
Dan: Sure, so what we are also seeing is that we are finding people that are kind of across the broad spectrum and that some of them are just starting to even know how to spell taxonomy. The joke is that it is confused with taxidermy, which I always like to throw in my presentations.
A lot of people are looking for a silver bullet and just like a technology solution guaranteed is not a perfect solution. But there is a lot more out there than most people realize that do help on the auto categorization front.
More than likely it's still going to be supplemented by some manual efforts. So anybody's efforts they have already put into doing manual taxonomies and theorize and anthologies and things like that will not be lost. It will just be more effectively leveraged.
Jeff: Yes, that is absolutely correct, Dan. We don't believe there is a silver bullet just to kind of throw at this problem that is represented by any one solution approach.
We think it is really an ensemble play and that auto categorization, as does enterprise search, has a very important role in the ecosystem relative to taxonomy. In fact we actually see almost a virtuous cycle with respect to taxonomy being used as an expression of the terminology, the definitions, the knowledge and know-how of an enterprise.
That is being used as a guiding framework to some extent to teach an auto categorization algorithm about what we already know about our information relationships. So those two used in conjunction with each other tend to create a much better descriptive set of indices, a much better descriptive set of search results that make investments in these platforms more leveraged and of more value for the typical enterprise.
Dan: Right. When we are doing our training or consulting we are always making sure that people are looking at whatever artifacts they have in place now because there has to be something that can help jump-start whatever effort they are taking.
There is no sense in just assuming that it is all garbage and throwing it right out the window. That was entirely a waste.
Jeff: Correct. I think in fact as we talk about the transition from Taxo 1.0 to Taxo 2.0 to 3.0 that is what we are so excited about in terms of the value proposition around business semantics management which taxonomy is at the heart of. We are really talking about leveraging existing investments.
This isn't about rip and replace. This is, we think, about solving for an overlooked piece of white space that when solved for will help companies better leverage the investments they have already made, and that is what I think is exciting for us in terms of where SchemaLogic is at in the market today.
Dan: I know we have talked a bit on previous occasions about the folksonomy model and how that is potentially at odds with a more traditional library sciences approach. I'm curious on your take on this.
People who have read any of my blog entries can tell I am pretty interested in the folksonomy idea. Unlike a lot of people, I don't think that is a silver bullet either, but there are some ways to meld the two and the way that SchemaLogic is actually providing the framework to support both of those is really interesting and relatively rare at this point.
Jeff: Yes, in fact you hit it right on the head, Dan. I think that as I was saying in my preamble that if one has a view that Taxo 1.0 is slow, brittle, hard to implement and kind of backward looking, a view of what has already happened, our vision of Taxo 2.0 really is to take those fairly static classifications of an enterprise or of an industry plus an enterprise that are expressed in taxonomy, not very frequently changed, and marry that with some of the populist approaches that we are seeing with respect to Wikis, folksonomies, tags and clouds.
What if you could provide a framework that underneath that overarching taxonomy where you do have a fairly static set of classifications, codifications, and information relationships? What if underneath that overarching structure that is a taxonomy you could really allow people to really collaborate on the evolution of that semantic, the evolution of that knowledge and know-how that a taxonomy represents in a folksonomy type of fashion?
What if you allowed your subject matter experts or the knowledge workers of the enterprise to participate in the broadening of this model or this common semantic underneath taxonomy?
That is where we think that we really now begin to talk about literally an up-to-the-minute definition of the semantic expressed in taxonomy for enterprise that can be used to information-enable key processes, to improve how search works or to improve how auto categorization and search works in the context of say a global portal or a global brand management.
That is where we think that leveraging some of these collaborative and participatory radical trust approaches to evolving information is very relevant with respect to how we look through our lens of taxonomy 2.0.
Dan: I think it's a pretty interesting time right now for this sort of world and one of the traditional cons of more library science traditional taxonomy approach is that it is heavy lifting to keep that thing up to date and that's why I don't think the phoxonomy approach is absolutely the way to go. But it does, between that and auto patherization and smarter natural language or whatever techniques that the ability of this body of work to have continual life rather than to become a dinosaur and impediment to work is actually a core business asset and can certainly propel you to the future.
Jeff: Yes, correct. In fact you bring up a good point, Dan. I think that if we were to do this as a triangle with respect to organizing principles the upper aspects of the triangle would be your more classical Taxo 1.0 where you have, as I mentioned before, fairly static and perhaps a slower changing evolution of the model and relationships between information elements in terms that taxonomy expresses.
But then in the lower regions of that model as you get closer to the broadest part of the tree, so to speak, this is where we see this opportunity for evolution.
So you have the guiding principles that taxonomy represents but then within those guiding principles you have the ability for kind of this socio-semantic evolution that we model in taxonomy and anthology which we think is a very powerful combination of leveraging the best of Taxonomy 1.0 with collaboration and architectures of participation which we think is a backdrop to Taxo 2.0.
Dan: Right. And we don't have the time, unfortunately, to talk about Web 2.0 and all that that entails although we are kind of hearkening to that right now. As far as Taxo 3.0, do you want to lay out what your vision is there?
Jeff: I would be happy to. We take this backdrop of Taxo 2.0 and effectively SchemaLogic's view is the evolution to Taxo 3.0 is effectively business semantics as a service.
What if we not only had a collaboration facility as we mentioned before but the ability for semantics, as expressed in taxonomy, to be subscribed to as a service by individuals to perhaps promote better search or vertical search subscribed to by groups, organizations, training partners.
What if everyone in my extended supply chain could subscribe to a common semantic therefore allowing me to optimize supply chain execution across 500 suppliers?
We think that this is again semantics as a service being a key enabler for information enabling, investments and platforms that big companies have already made.
Enterprise search, content records documents management, portal access, federated access to information that we think this plays a role in helping to manage. So of course this technical jargon is only as good as our ability to map at the business impacting our lives, which is really what matters for the enterprise and is really very much the mission of SchemaLogic in terms of promoting the business benefit of why this matters to the big companies.
Dan: Yes, this isn't just some academic exercise or another excuse to come out and buy technology just for kicks. This is about trying to make business a lot more effective and it's been talked about for years.
But silos still exist and there are some good reasons for them to exist and some not so good reasons. This is one way to be able to cut across the silos and actually unify them for real. We saw that with portals, with search and with federated search in particular. There is just a whole lot that is going on here that people are aware of and hopefully between the two of our companies and many others we will get that message out to the market.
As for ROI, I could rant and I have all day long about the dangers of the ROI rat hole, soft dollars and hard dollars and all that if you do need to justify your investments, which in this day and age I can't imagine anyone not having to apply some real ROI to your project.
But when it all comes down to it you either believe that this gives you a competitive advantage basically or you don't, in which case I'm not sure what you are going to be doing with yourself in several years with your information.
This really can impact the ROI of your organization.
Jeff: Yes, that is correct, Dan. As I've mentioned we believe that our solution approach with collaboration of participation at the core allows companies what we describe as information enabled key processes such as Web 2.0, B2B or B2C enablement thereby opening up new revenue channels because I can optimize my demand chain and create access paths to sell business services that otherwise would be hard to find if it weren't for the internet.
Or for facilitating global employee or customer self-help, self-service via optimized information access through portals. Or better dealings with regulatory oversight for all information in regulated industries with high degrees of terminology varyings over time which is very much the case in the bio farm industry as an example.
So looking back to the notions of Taxo 1.0, 2.0 and 3.0, it's my belief that with this notion of Taxo 2.0 and 3.0 we are really arming classical taxonomists with the means to finally make business impacting “why” arguments that will support spending in ongoing investments that we believe have oft been overlooked or simply not made as no one really understood what Taxo was or why it mattered.
I think this brings it's role much more into the mainstream as a check box item that every, as we call it, Global 5000 company will have to address, which is this notion of business semantics management and the role it plays in enabling and leveraging competitive advantage through better information management for the enterprise.
Dan: It's been interesting. I've been with Delphi for almost 12 years here and there has definitely been a revolution in the last several years on computational linguistics and a whole lot of other things around taxonomy, auto categorizations, search, portals and obviously the web surrounding all this.
There has been a ton of market shifts on the supplier side with the merger of autonomy and verity. IBM all their labs and acquisitions they have made in the last year. Oracle is doing the same and a whole lot of startups, everything from day zero startups to the older startups.
It's a really interesting time right now. It's well worth people looking at this and seeing what can these things do for you. And be very aware, too, of what they can't do for you.
There is still work involved but there is some honest to God benefits to this once you start to roll it out and ripple it throughout your systems.
If people listening or reading would like any further information, feel free to contact us at the Delphi Group at www.DelphiGroup.com or 617-247-1511 or anyone at SchemaLogic which is www.SchemaLogic.com or 425-885-9695 if you should choose to call.
I want to thank you for your time, Jeff, and thanks to everyone out there who is listening or reading. Any comments that people have feel free to post comments to any of the blog entries that would be surrounding this. Otherwise get in touch with us directly.
Jeff, do you have anything on your part?
Jeff: Great, Dan. Certainly I want to thank you for taking the time in hosting this session as well as Delphi's ongoing commitment to providing thought leadership and sponsorship of what we think is an exciting part of solutions that provide competitive advantage.
Of course in closing, from a SchemaLogic standpoint, please feel free to get our web site or talk to us directly in terms of providing more definition around the notions of Taxonomy 2.0 and 3.0 as we are very happy to talk about how we help companies think about building these ROI's to support what we think is an exciting solution to integrate into how companies manage information today.
Dan: Thanks again, Jeff, and thanks everybody else for listening/reading.
Jeff: Great. Thank you.
Comments