Semantech Ensights

What should the top priorites be for our nation's first CTO?

2008-11-13T09:01:00.000-08:00

The role of national CTO will provide the Obama administration with the first true technology transformation capability at the cabinet level. This is a very important consideration, as previous approaches have inevitably gotten lost within organizational 'stove-pipes' and competing interests.

The most important thing that the new CTO can do is help to define national priorities and set the roadmap expectations for where we need to be in the next 5 or 10 years. This exercise has two parts:

1 - Setting expectations for technology-focused national goals.

2 - Setting expectations for how technology can help resolve other national issues or problems that may not necessarily be considered within the domain of the CTO.

Specific priorities for either of these categories might include the following considerations:

1 - Adoption of Intelligent Healthcare practices (i.e. this extends much further than the adoption of EMR/EHR systems - Intelligent Healthcare post ).

2 - The ability to track policy initiatives in an automated fashion and link that also to program performance tracking.

3 - Focus on "Technology Multipliers" - i.e. technology-based initiatives that are specifically designed to solve more than one problem at a time.

4 - Re-evaluation and reinvention of traditional industries (already perhaps underway in Energy and Transportation - so it could focus on how technology facilitates and accelerates this process).

5 - Empowering citizen groups and government entities (state, local, federal) to help solve problems through "Federated Collaboration."

While some of these activities have been occurring for a few years, the strategic role of the national CTO and the commitment by the Obama administration could provide the momentum to take such efforts to an entirely new level.

Copyright 2008, Semantech Inc.

Is there a Cultural Bias behind Innovation?

2008-11-08T05:31:00.000-08:00

We'll start with an interesting and perhaps controversial question - there is a widespread perception that certain cultures are inherently more innovative than others, is this myth or reality? This question strikes to the heart of whether innovation can be instilled, taught or otherwise coaxed from sources where it currently doesn't exist.

The traditional thinking goes that the United States is a leader in Innovation and other markets / regions such as China, Japan or India follow in our footsteps by becoming more effective in application of already existing technologies / innovations. This viewpoint of course only seems to hold water when viewing these regions at our present time (in other words, this doesn't represent long-standing historical precedents).

So, let's assume for a moment that the United States is more Innovative in terms of a few key indicators:

Number of patents issued
Number of scientific papers published
Number of academic institutions engaged in research
per capita income of nation targeted towards to R & D
Number of industries started or more or less headquartered / centrally focused in the country

So, if we follow the assumption, why does the United States exhibit such advantages? Is it a result of our previous focus on Cold War research and the Space Race? Is it a byproduct of our economic climate, or is it somehow a product of our culture ? Or can these elements readily be separated ?

The reason I'm asking this question today is due to a great deal of discussion that I've overhead lately regarding the overall decline in enrollment within the United States in Science and Technology degree programs. There seemed to be a lot of concern that we are losing a competitive advantage as a result of this situation. It got me to thinking what factors might be involved in determining how or why this type of success occurs and whether it was truly related to the numbers of folks attaining science or tech degrees.

The short answer in my opinion, is yes, there is a cultural bias. This has nothing to do with ethnocentrism or jingoism - but rather a recognition of certain factors which are normally not considered. I'll address what those factors are in the next post...

Copyright 2008, Semantech Inc.

The Tools of Innovation part 1: Concept Maps

2008-11-08T05:29:00.000-08:00

If we wish to view Innovation as something that truly can be reproduced or engineered (at the individual or organizational level), then there must be a set of tools to facilitate those activities, right? It is perhaps hard to imagine what they might be without ever having worked with them and of course that's exactly how I approached the subject just a few years ago.

But I started small and worked my way up and all that while a new breed of tools were being introduced to the market as well. My introduction to this realm was the "Concept Map." Concept mapping tools are akin to Mind Mapping tools but without some of the constraints (in terms of how thought trees are captured and displayed).

The wikipedia definition for Concept Mapping is - Concept mapping is a technique for visualizing the relationships among different concepts. A concept map is a diagram showing the relationships among concepts. Concepts are connected with labeled arrows, in a downward-branching hierarchical structure. The relationship between concepts is articulated in linking phrases, e.g., "gives rise to", "results in", "is required by," or "contributes to".

In other words, the concept map allows you to quickly build, visualize and capture a crude ontology of any concept/idea. The tool that I've been using is freeware, it's the de facto standard - CmapTools. The product is in transition and it is now being sold commercially as well as Ceryph Insight.

What I like about it is the ability to save your designs to XML and if you get the plugin (Cmap Ontology Editor), to save your Concepts Maps as OWL files. So how does one use a Concept Mapping tool? Here's how I do it:

Step 1: I have it loaded on a laptop and ensure that there is a projector available
Step 2: I have an agenda of related topics, issues, assumptions or ideas already gathered in bullet point fashion on a sheet, which is either for me or handed out to a group.
Step 3: Throw it on the chart - basically one begins to place the various elements on a screen and see how they relate to one another or whether they have merit.
Step 4: Downselect - focus areas are developed and dead-ends are deleted.
Step 5: Pretty it up - This is where we add some design elements to make the diagram outputs suitable for inclusion on presentations, web sites etc. This is also when you decide what text to add to the connecting arrows...

So what does it look like when you're done, well, it can look like any diagram, here are some examples:

This diagram follows a more traditional concept map portrayal

This diagram looks less like a concept map but can still be saved as XML

Copyright 2008, Semantech Inc.

Metaphors for Design

2008-11-08T05:28:00.000-08:00

My all time favorite example of the use of metaphor goes back to the middle ages; traveling bards used to present stories and poems to audiences from town to town before the advent of print. This was an interesting variation of storytelling tradition that occurs in most cultures – but what many of the bards did to manage their storytelling experience was construct the story-telling experience out of a metaphor. The metaphor was often imagining a house – and within each room the storyteller would locate an important part of the plot, items within the room would trigger specific details in the retelling. In many ways what this represented was a mental discovery and / or content management application. We’ve all variations of this them today – the most obvious example is what we refer to as a “Chat Room,” but other applications have used the room and structures metaphors for similar functionality.

So why metaphors, what’s the value proposition we’re describing here? A metaphor is generally only useful if it applied as a means to leverage existing knowledge to help facilitate some other action. That action might be story telling or driving heavy equipment, but in most cases the action we’re hoping to facilitate is somewhat complex in nature. So the question is; can someone learn everything they need to know about fulfilling this new complex action in a reasonable timeframe, and if the answer is no, then what exactly can be done to reduce the learning curve? Simply because someone is using a new technology to accomplish a familiar task it doesn’t mean that all aspect of the task need to change also. There are many other benefits to this type of continuity besides a reduction in learning costs; they include:

Greater likelihood of acceptance and adoption of the capability

A Greater likelihood that the new capability will be able inter-operate with examples of the old capability still in place elsewhere

The ability to improve on but maintain process approaches that have proven themselves to be effective.

Example

The best example of something I’ve been involved with that is essentially a metaphor is an online magazine or newspaper. Perhaps this was an easy step to take due to the nature of Internet and it’s ability to disseminate information, but the fact of the matter is that it allowed for massive restructuring of the previous business model (including operations such as printing and delivery) without significantly changed the user perception of the capability or product. The biggest difference of course is the lack of paper involved with the Internet versions; however this is something that is being worked out by vendors right now – new delivery interfaces that are nearly as thin as paper and are bendable will be available in the next several years. Along with anti-glare screens this will come fairly close to the hands-on tactile experience of reading a real magazine (and much more efficiently as the device can upload any magazine’s content through wireless connection and is not disposable).

Following the metaphor then, all aspects of content design (including advertising) would approximately the current look and feel of typical magazines. The one area that the current manifestation of the metaphor cannot handle which the new one can is the ability to embed interactivity with other applications. For the interactivity, the only guidepost we have with current publications are written games such as crossword puzzles. In these types of situations though, I think what actually happens though is that we tend to borrow from multiple metaphors to build a new capability. So, the new magazine borrows metaphors from video gaming, broadcasting, old publishing models and B2B web applications to achieve a new type of approach. This is not too different than what has just happened with the new i-Phones.

Ideation

2008-11-08T05:26:00.000-08:00

This is a new term that has come in vogue over the past several years and one fairly closely aligned with the notion of Innovation engineering. I thought it might be worthwhile to do some comparison and contrast between the two concepts (with the obvious disclaimer up front that I don't believe them to be the same thing).

First let's assess the definitions:

1 - Ideation: and this my take on it, as I somewhat doubt that there is any industry standard view as to what it is just yet. Ideation represents the front-end portion of concept and product development efforts. It is somewhat synonymous with brainstorming but more comprehensive than that. It is an attempt to standardize the creative process early in the initiative stage in order to produce more consistent or agreeable results. It is sometimes loosely coupled with follow-on product development efforts but seldom fully integrated with them. The connection though is why many PLM (product lifecycle management) vendors refer to ideation.

2 - Innovation Engineering: This represents an undercurrent or element within a larger set of lifecycle processes. While there is always a focus on innovation at the front of any endeavor that is not where innovation ends. There is a bit of a dual meaning at play here as well, when we talk of innovation engineering were not only referring to the application of innovative thinking to other problem sets, we are also discussing how to manage the creative process of innovation itself to make it more predictable.

So, we can see that there are some obvious overlaps here, including:

1 - The focus on creative problem solving
2 - The ability to produce more predictable results
3 - The emphasis on applying creative perspectives at the initiation of projects

The differences could be characterized this way:

1 - Ideation is more front-loaded whereas innovation engineering seeks cradle to grave improvements
2 - Innovation engineering is meant to have more of an applied focus
3 - Ultimately, ideation will never provide substantial metrics to determine its effectiveness given its esoteric and usually disconnected nature (i.e. lack of integration into the follow-through efforts)

copyright 2008, Semantech Inc.

How do you choose a right architecture?

2008-11-06T14:51:00.000-08:00

"Right" is always subjective and by that I mean it is driven by the circumstances of your engagement:

* The Client (systems / NETOPS) Environment
* The Client Organization
* Functional Requirements
* End User Expectations

I'm assuming that you're referring to a solutions architecture - this can be captured at multiple levels using any number of frameworks or notation conventions. Usually the framework, notation and tools you use will be dictated by the client and if not obviously you should use whatever you're most familiar with. The architecture often also has a temporal aspect to it, in that many clients will want to better understand their current situation, the "to be" solution and the transition from 'a' to 'b'.

The design choices are tied directly to the unique problems and tradeoffs that exist within the target organization. So, for example if a client is having major difficulties with performance on database systems and perhaps their SOA services you would look at using caches, XML appliances and more efficient messaging.

There is no one size fits all scenario. As much as we'd like to capture all of this with patterns or COTS software there will always by unique requirements. The key is to understand how much of the time-saving techniques and tools can be applied to the enterprise problem space along with the client-specific architecture / design.

What does the term "Personalized Medicine" mean to you and in what way do you think it has the greatest potential?

2008-11-06T14:48:00.000-08:00

I think a better term might be "Contextualized Healthcare," this is more sweeping and covers a lot of converging trends (as well as technical capabilities). Medical care is being handled in a more pragmatic way than before through an improved understanding of both personal and group level contexts - this covers as you say both prognostics and diagnostics but also provides us a framework for managing continuity of care for individuals across their lifetime and for various demographic groups.

The age old dilemma for Medicine and Philosophy is balancing a priori and a posteriori knowledge in practical application - the one mitigating force here is "context." As we become more focused on the ability to understand and predict specific scenarios (rather than dependence on generalized theory), quality of care will improve dramatically.

Copyright 2008, Stephen Lahanas

Semantic use of H1 heading

2008-11-06T14:46:00.000-08:00

As with most things related to Semantics, context is everything. Back in the beginning days of the web, H1 was the site title and sometimes even served as a logo. Today H1 is more akin to a major heading, however it was never quite large enough (without modification) to serve as a true headline in the Newspaper sense.

Now that we do have much more flexibility to modify tags using CSS, we can in essence define our own semantics. The other side of the question is how does the search engine world view semantic extraction of meaningful information from sites. Well, it is easy to develop algorithms or business rules that assign importance based upon header designation and placement.

Long story short - the "should" aspect of the question is perhaps a bit too prescriptive. What we ought to be doing, always, is to continue to explore new ways to improve exploitation of this technology - both from provider and consumer perspectives.

copyright 2008, Stephen Lahanas

IT Talent Shortage: Should IT executives do more to create local talent instead of just complaining about the trend?

2008-11-06T14:44:00.000-08:00

There is no IT talent shortage and never has been. This myth was and still remains a ploy to promote Globalization. Millions of IT workers have had to switch careers here in the states due to the odd fluctuations caused by shifts in policy related to outsourcing and visa approvals.

We should of course always encourage local development of talent, but it is important that they understand that the career path they've chosen is not subject to exportation elsewhere...

copyright 2008, Stephen Lahanas

What aspects are important to have in a "online newsroom" (social media newsroom)?

2008-11-06T14:43:00.000-08:00

Well, the RSS is important (and other social media tools as well), but I think it's worth keeping in mind the scale of the task that you're tackling with an online newsroom.

It is more or less the equivalent of a "news wire," in other words you are more than likely going to be providing a continuous stream of content. The most important thing for achieving that kind of output is securing relationships with a number of content providers. So, what is really necessary is a platform that supports multiple author contributions w/ minimal editorial management from the publisher...

Copyright 2008, Stephen Lahanas

Can Artificial Intelligence support ad hoc decisions?

2008-11-06T14:41:00.000-08:00

If the question is "can this occur potentially," then the answer is yes, without question. Is it occuring effectively now, I doubt it. The key here is the ability to identify enough potential "paths" that later qualify as ad hoc decisions.

In other words, all of us through experience build up an intuitive inventory of potential actions and thoughts that could be drawn upon as specific responses to various scenarios. Once you reach a critical mass of such related knowledge, an AI machine / program could more or less mimic a human ad hoc process.

Last Fall I was at a NASA event and they brought out a NOAA robot that did a little of this, mainly to entertain the kids at the event - and it was impressive, however the knowledgebase / engine needed drive something like that really needs to be a 1000x larger...

Copyright 2008, Stephen Lahanas

Will the Internet help bring about enlightened societies that actually think critically?

2008-11-06T14:40:00.000-08:00

Yes, in fact it is fairly clear that this is occurring already. Let me caveat that a bit though, I believe that folks within societies have always had the ability to think critically and collaborate in meaningful ways, however the structures within most societies did not support those abilities.

This is an important distinction - we're not creating a new capability, we're facilitating an existing one. So we're moving from one-way information dissemination to nearly infinite communication paths - if we view enlightenment not as the sum of all cultural knowledge but as the shared potential of people working together without limitations then yes we are definitely becoming more enlightened.

Copyright 2008, Stephen Lahanas

Is Radio & TV Advertising Dead?

2008-09-09T13:01:00.000-07:00

It seems to me that it is not still alive but it is in effect merging with the emerging forms of "IP advertising." This is going to become more apparent over the next five years but the (marketing) channels are blending and once unified it will become easier for clients to pick and choose coordinated campaign strategies. So what we’re talking covers a few basic principles:

1 – The dissemination mediums don’t necessarily affect the skills and practices involved with producing video and audio content (other than that the underlying production technology has also become digitized making it thus easier to transform content for multiple delivery scenarios).

2 – Advertising dollars are diversifying (especially into IP) but are also still expanding as a whole. The obvious growth area for IP advertising (the largest growth sector) will be the ability to bring it up to the same standards associated with traditional venues. The easiest way to do that of course is to produce content for all venues at once.

3 – Neither the venue or technology ultimately determines what type of ads succeed. There is still a ton of money being directed at really mindless advertising (in all delivery mediums). The greatest opportunities (in all market segments) will come for those who continue to approach the audience with respect and apply wit to their task.

copyright 2008, Semantech Inc.

Have you recently tried to hire a data quality team member?

2008-08-20T08:21:00.000-07:00

What have been your experiences of trying to find data quality skills for recruitment to your organization? Easy? Challenging? Complete disaster?

The immediate thing that comes to mind is - how do you characterize the skill set of a data quality professional?

* is it focused on data modeling?
* is it focused on data interfaces, translation, transport, validation?
* Is is focused on the COTS toolsets and if so at what layer in the architecture? (DBMS, ETL, ESB, BI)?

The success rate will likely have something to do with how specifically you answer those questions and how you then apply that to the search...

What does Web 2.0 mean to you ?

2008-08-18T15:00:00.000-07:00

And how does Web 2.0 provide Enterprise Level advantage ?

As far as Buzz-terms go, "Web 2.0" and "Enterprise 2.0" rank as the most confusing and least precise I've seen over the course of my career. This is becoming worse as some folks are adding "3.0" for different variations. Having said that I don't know if I have a viable alternative.

The use of collaboration technology on the web, while now in full swing, has been around at a substantial level for more than a decade - I just can't see that as new per se. Granted, many of the standards released by the W3C in the later 90s and early 2000's are allowing for much more functionality, I'm not sure whether we're seeing truly novel innovation or just "expected evolution."

When you mention Rich interfaces I couldn't help thinking about DHTML and the first wave of Flash (actionscript) applications. The technologies are of course different but some of the user-facing dynamics were the same. So, is the web really different now, different enough to deserve the 2.0 moniker?

And how do we quantify different; the user experience,
the depth of community activity or the web developer's perspective in terms of toolkits and standards to choose from ? And does the Semantic Web fit within the Web 2.0 concept domain ?

Open questions...

What Business Intelligence trends have you observed?

2008-08-18T14:58:00.000-07:00

In addition to any trends you are observing, which ones do you feel will make the most positive impact for business users of BI?

There have been some positive trends, although the road towards next generation or "Agile BI" has definitely been a slow one. So what constitutes Agile Business Intelligence?

1 - A greater connection with and allegiance to the end user. In other words co-opting the users within the development process. The democratization of BI - this refers to a more open, less costly licensing approach which recognizes that BI is applicable across the enterprise, not just in the hands a select few.

2 - BI is getting faster, provides more value and more is getting more cost effective.

3 - Greater flexibility in terms of including ad hoc reporting capabilities, coordination with unstructured data sources and user collaboration.

Perhaps the best way to bundle these considerations into a single definable trend is to view BI as moving to a more response, rapid and cost effective part of the enterprise.

How your firm match IT roles to ITIL roles?

2008-08-18T14:57:00.000-07:00

ITIL® (the IT Infrastructure Library ® ) is the most widely accepted approach to IT service management in the world. Many corporations today are employing these practices. How does your firm match IT roles to ITIL roles?

One way to think about it is this, ITIL represents a best practice approach towards standardizing IT methodologies and practices - the most obvious starting point to doing is aligning the Semantics of IT. ITIL represents at its core a set of generic or upper level Ontologies. Like any ontology though this one is useful only in the context of some type of consensus within the communities of practice that adopt them.

The Federal Enterprise Architecture framework has very similar ontologies in place and of course if you look at any number of large task order contracts, the set of available job descriptions (labor categories) are somewhat similar. But the use of job descriptions is different enough to make things confusing - more importantly the nature of the jobs themselves are changing now more rapidly than in previous decades.

What ITIL might be able to help with is to provide a mapping reference point or baseline against other competing or emerging job description ontologies.

What is your definition of a maturity model?

2008-08-18T14:55:00.000-07:00

"It is well understood how SEI/CMM defines maturity models, but what do you believe a maturity model should contain? OWASP is working on defining a maturity model around application security and is seeking contributors."

I'm a little bit skeptical that there even should be a maturity model for application development specifically geared towards security. How does one separate the security of an application from the application design itself? The maturity model focus ought to be comprehensive and inclusive or security from the application perspective.

Why do I say this? Well, consider this - what if the security maturity model (SMM) leads to greater separation of design and development roles (and compartmentalization between portions of the architecture) and perhaps then we will see arbitrary decisions about security being made without consideration to application performance with no ability to reconcile between the competing interests?

Using social networking tools for knowledge management

2008-08-18T14:52:00.001-07:00

question:

"Hi. I have been studying role of social and professional networking process and tools (such as Linkedin, Ecademy, Twitter, Myspace, Facebook etc) in knowledge management. More specifically, aim to find out (from your experience) whether these tools can be used for knowledge transfer and greater learning amongst employees, teams, special interest groups, community of practices, etc. I'd like to obtain your insights on this."

Yes, and more importantly these capabilities provide an excellent interactive learning platform. In my view, the primary reason why Knowledge Management systems in general have not met their potential is due to the fact KM only views half of a larger integrated process. We exploit knowledge to learn - KM as archive without a focus on learning is empty...

So, the social networking, Web 2.0 technologies not only provide a wonderful source of unstructured data / knowledge they provide a significant portion of the learning environment necessary to exploit KM.

Linkedin Question "SaaS vs. SOA"

2008-08-12T08:43:00.000-07:00

Question:

[ SOA and SaaS, Is this the current IT battleground?

There are strong arguments being made in both of these critical areas of IT. Each have their proponents of how the model can best deliver value to the business world, with compelling arguments being made in each area. Do you feel that the two are mutually exclusive or whether there are possibilities for a mixed architecture?

Whatever approach is adopted data mastering is still a major challenge - how would you ensure that the Customer master (and other masters for that matter) remain clean, up-to-date, and duplicate free?

With a SaaS approach how do you propose a) protecting corporate data, and b) making the data available for Business Intelligence applications? ]

response:

I suppose that in a smaller enterprise these two approaches or options might be considered mutually exclusive, especially if there are only a handful of mission critical applications involved. It is important to understand though that neither of these approaches necessarily represent a holistic architecture by themselves.

The other questions regarding Master Data Management & BI apps can be approached several different ways:

1 - The MDM capability might well be another SaaS service option. (security will always be an issue regardless of where the logic & data is located)

2 - MDM functionality and sensitive analytical functions can remain in-house (and merely pull data across from the SaaS data source).

3 - It is more than likely most enterprises will settle upon mixed architectures.

How do you see Knowledge Management defining the current state of research in life sciences?

2008-07-22T08:31:00.000-07:00

(question response:

I wish to focus on aspects relating to intelligible connection between databases, data profiling interfaces, pathway design systems... What would be the future of new information management systems when handling huge amounts of disparate data?)

Well, what you're describing goes somewhat beyond what is traditionally referred to as Knowledge Management and it can also be abstracted somewhat from the industry & problem sets you're describing. What we're discussing is a wider context of data architecture which necessarily must encompass both structured and unstructured data utilizing fusion and federated management. The foundation for this is a flexible set of semantic models that provides the necessary abstractions between the other architecture layers and constructs.

In the DoD or DHS we often build fusion engines with Common Operating Pictures at the apex of the solution - in many cases the primary interface includes visualizations (GIS focused, i.e. map based). The core problem is the same in that there is a need to cross correlate and manage massive amounts of data / information but also to do that in the context of specific user definable scenarios. So the answer to the question is essentially yes, the capabilities available to you will make a profound difference in what you can or can't do with your research.

What project management tools do you like and why?

2008-07-22T08:19:00.000-07:00

What you're talking about goes beyond project management or project management software - the issue is what type of development & sustainment lifecycle management processes you have or would like to have and then how you can automate them. Project management software such as MS Project does not allow you to go into detailed requirements or really facilitate development processes - basically you're just tracking schedule and resources (the cost tracking in not well development in Project).

Those elements are important but provide an incomplete picture. There are a number of vendors who have either Product Lifecycle Management tools or PPM (portfolio & project management) tools that come close to capturing the big picture but ultimately your organization needs to choose its preferred process (formal or Agile or some hybrid) and then choose the appropriate software which facilitates that software. As an example, I went through this process several times on my own in the context of supporting several large clients and I've built a practice around what I developed (see links) - basically I came up with a meta-lifecycle process (to support multiple project teams) and employed a modifed version of a Product Lifecycle Management tool to automate it. The software I use most often is Accept 360, but there are a number of choices out there and the one that makes sense for you will become clearer once you decide how you intend to manage your software development strategically...”

How to get start with SOA (Service Oriented Architecture)?

2008-07-22T08:17:00.000-07:00

(question & response posted previously on Linkedin.com)

The one thing I'd mention is that there can be wildly different expectations regarding what it is you may wish to accomplish with SOA; if for example you are merely looking to port some legacy applications over to web services - that would involve an entirely different set of activities than say if you wanted to transform your enterprise using SOA as a foundation or catalyst. I tend to work in the latter space - and as several have noted already in response to your question, there is more to SOA than SOA - no, that's not a riddle, it's just a recognition that if your scope is wider then you must consider more than just the application architecture and business governance typically associated with SOA solutions. I tend to refer to this as Service Oriented Integration... (which includes other considerations such as Semantic Integration).

Will SaaS change the face of Enterprise Architecture and SOA ?

2008-07-22T08:15:00.000-07:00

(question & response from Linkedin.com)

I don't think it will have much impact on EA given that the primary role of EA is to define the enterprise in whatever form it may happen to take. EA and EA frameworks must support, out of necessity, a significant level of flexibility in order to remain relevant and useful as a tool. SaaS is becoming the SOA 'COTS' - i.e. the customer facing 'out of the box' capabilities as opposed to SOA stacks that power custom-built / coded enterprises.

Viewed in that context, the emergence of SaaS is perfectly logical. The impact to most enterprises will come in the form of near-term interoperability issues - this though should pass as many organizations may choose to bypass management of their own SOA stacks (for a lot of reasons) and the SOA vendors will provide more opportunities to run 'foreign' service logic and/or integrate with them through mash-up type interfaces.”

Team Performance

2008-04-11T11:50:00.000-07:00

I think that this topic is one of the more relevant considerations regarding potential project success as it implies a good deal of specific areas for improvement that can be applied in nearly every workplace. I’m not sure that team performance is dependent on leadership – I’d refer to it as facilitation. A facilitator enables the group to lead itself in a more democratic fashion perhaps than if it were being directed by someone who had specific authority over the group. I’m not sure I fully agree with the group dynamic lifecycle concept - That concept lists out the following stages that tend to occur in most small group scenarios:

· Orientation

· Conflict

· Brainstorming

· Emergence

· Reinforcement

The second step in the process, ‘Conflict,’ is out of place here or perhaps merely mislabeled. I think that differences in opinion should be sought, but not in the context of any sort of conflict. The premise for building the group should work through any potential conflicts before its inception and thus help provide a more realistic or specific foundation for the group to start upon. A group that is still largely unclear regarding its purpose is very likely to fail in its mission – I have seen this happen many times.

The best example that comes to mind right now is the network of groups being used across DoD right to build data vocabularies and ontologies. These groups are referred to as Communities of Interest (COIs), DoD CIO (2007). Many if not most of these groups have failed thusfar to achieve their original objectives. Conflict has played a role in this but also the difficulties in finding representation and coordinating those resources has made its mission difficult. Small groups are hard enough to manage well, building in conflict as an expected part of the lifecycle seems to me at least to be counterintuitive.

Project and Organizational ROI

2008-03-27T17:13:00.000-07:00

Return on Investment (ROI) calculations for Information Technology projects are much more subjective than traditional organizational finance. Traditional organizations deal with a smaller set of variables and often are focused on very tangible management of physical assets or products or services. What makes IT ROI more difficult to calculate is not only the higher levels of risk associated with it but also the difficulty in assessing expected versus actual benefits of automation or access to improved information. For example, if a decision support project is being developed to give key executives real-time oversight of all business activities, then it stands to reason that they might make more timely decisions that allow them to mitigate problems sooner and exploit opportunities more rapidly. However it might not.

There is no guarantee that the system will produce any beneficial results; the only thing that is certain in this scenario is that a certain amount of funding is being expended to build or buy the system and more will be required to maintain it. What’s worse, is that if the company does seem to do better sometime after the system is deployed it will not always be easy to assess specific performance or productivity gains to decisions made based upon availability of the new system. For many systems like this example, it is nearly impossible to prove the true ROI of the project and thus the entire ROI process becomes more of marketing pitch than financial analysis. This situation gets even more complicated in non-profit environments, such as the federal government, where no profit can be derived from any IT effort. All that can be projected in those environments is cost savings or capabilities enhancements.

In fact, from a government perspective, many folks have tried to go beyond ROI measurements to determine value, CIO Council (2002):

An ROI calculation for a particular investment may, in fact, be negative, but the intangible benefits such as customer satisfaction and easy access to information, may justify the expense. ROI is an important metric, but it needs to be balanced with a rigorous analysis of all the value factors.

I also recall very clearly in years past that nearly every IT Hype Cycle has been closely associated with an ROI blueprint, showing how much money would be saved if the new technology were adopted. In most cases, these blueprints were self-serving and an inaccurate. This led to skewed expectations and in my opinion an early exit for many of these trends. To some extent this is also what happened to E-learning (the medium we’re using now); that technology is only now recovering from the original hype ROI assertions.

Does Modeling have a place in Evolutionary Design?

2008-03-24T11:34:00.000-07:00

Any field of practice represents a community of consensus; in other words, many of the core elements of the practice are based upon shared beliefs rather than scientifically proven principles. These communities often go through cycles of orthodoxy during which certain belief or techniques are favored and others are not. These represent preference trends more than anything else. The only way to discern the true viability of a particular technique or idea is to track it across time and through specific implementation examples.

Having said that, evolutionary design as a formalized practice area is still somewhat of a trend. Many might contend that it existed as an informal practice since the beginning of the IT era, however we must view it in the context which it is currently being presented. Evolutionary design posits that accurate design is unlikely or difficult to complete in the initial phases of a lifecycle and rather than trying to achieve something that is likely impossible it is better to parse the design process out across the larger implementation lifecycle.

In many ways is it a bit of an extreme reaction to extrapolate that simply because the initial design isn’t 100% accurate that an up-front design effort is not worth the effort. On the contrary, even if the initial design is only being used as strawman, it serves and important purpose in helping to organize the thinking of the group. The real question (which has already been brought up in this question thread) is how to correlate an evolving code design with the models which inspired it or represent it. The answer to this question is subjective and based upon the composition of the team supporting the project. If the team has both designers and developers on it then being able to keep the models and the code in sync would not be as difficult then if the entire team consisted of developers. Naturally, people who are trained to do a thing well tend to stick to that thing which they do well – developers develop, writers write and modelers model.

So, each situation is unique and potentially dynamic. A reasonable approach that can be applied in this scenario is this:

The initial model can be developed in a prototype fashion (i.e. it is built to be updated with much detail left incomplete at first to be filled in later).
The model updates can be associated with design review exercises – as the team gets back together details are added to the model as insights are captured in the real time meeting. This too is not meant to be comprehensive, but allows the model to generally keep pace with development changes or additions. Tracking this also allows the team an easy way to assess the scope of changes that have occurred from meeting to meeting as each version of the model is captured separately.
As elements (units) of the development are completed more formalized details can be added to the model if there is an intention to take the project beyond prototyping, if not then there is no need to document the design.

Now that I think about this question, it might be worthwhile to step out of our immediate context for a moment and explore an analogy. When someone is writing a script or a book there are two approaches to take. You can either let the characters take you wherever your imagination leads you or you can plot out the story in advance. The script writing process is very similar to what we’re discussing here – a group of writers start with something (a pitch, a sample, a draft script), and collaborate on the script until it takes final form. Sometimes the script writing includes storyboarding, sometimes the storyboarding occurs after the script is largely completed.

The expectation is always that there will be an evolution of the content as it moves through the process; however that does not in any way preclude attempts to provide comprehensive planning up front. I mention this because as a practice area, script-writing has a much longer history than evolutionary programming. Script writing seems to have settled into a more flexible and reasonably pragmatic view of the development process.

Exploratory Testing & Agile Methodology

2008-03-22T07:17:00.001-07:00

Exploratory testing is most often considered a domain of usability testing but could easily be extended to cover just about any testing situation. The thing I like about exploratory testing is it sheer flexibility. It is for this reason that I tend to include it as part of an Agile approach or methodology.

The primary difference between exploratory testing and more formal testing paradigms is how the scenarios or test cases are derived up front (or not). In other words, Exploratory Testing begins with broad strokes rather than intricate detail - it allows us to look for major issues and help define the tests as we're going. While we're exploring we look for core functionality and stability - if we're more focused on site design, we take assessments of comprehensive aesthetics as well. As we move further into the test we can begin providing more classic metrics regarding performance and so forth, but then again exploratory testing often is not the only testing that is done - the determination of what tests cover what metric categories or concerns is very flexible.

Copyright 2008, Semantech Inc.

Business Analyst Salaries

2008-03-19T06:04:00.000-07:00

How Much does a Business Analyst make in the US ?

It really depends on the nature of what you're asked to do. If you are a B.A. and an architect and help them to reengineer workflows or business processes using BPEL or BPMN and are familair with process methodologies (perhaps ISO CMMi) then you'll tend to be priced on the high end of the scale. Most well compensated BAs tend to have the following skills in their portfolio: 1 - Requirements analysis & engineering 2 - Business architecture (conceptual data models, domain models, user cases). 3 - Process Engineering (as noted previously) 4 - Presentation and report writing skills 5 - Financial analysis (EVM, business cases, work breakdown estimates etc.) As with most professions, the more you have a chance to learn on the job the more likely you'll be able to move up the ladder.

copyright 2008, Semantech Inc.

Taxonomy Seminars

2008-03-19T06:01:00.001-07:00

I'm looking for a good course or seminar to learn about taxonomies; building an implementing one for an online KM-based business library. Suggestions or recommendations?

Taxonomies are fairly common sense endeavors, the nature of your environment will drive the structure, so for example if your company focuses on natural resources it will include those parts of existing geologic taxonomies that already exist and then be mixed with taxonomies that corrleate to the types of technologies you use (in connection with them). The best way to do it is just to give it a try, using vocabulary or ontology tools available for free on the web. I always start with mind maps or concept maps and then move to other more complex tools later...

Copyright 2008, Semantech Inc.

Knowledge Management & Life Sciences

2008-03-19T05:57:00.000-07:00

How do you see Knowledge Management defining the current state of research in life sciences?

Well, what you're describing goes somewhat beyond what is traditionally referred to as Knowledge Management and it can also be abstracted somewhat from the industry & problem sets you're describing. What we're discussing is a wider context of data architecture which necessarily must encompass both structured and unstructured data utilizing fusion and federated management. The foundation for this is a flexible set of semantic models that provides the necessary abstractions between the other architecture layers and constructs.

In the DoD or DHS we often build fusion engines with Common Operating Pictures at the apex of the solution - in many cases the primary interface includes visualizations (GIS focused, i.e. map based). The core problem is the same in that there is a need to cross correlate and manage massive amounts of data / information but also to do that in the context of specific user definable scenarios. So the answer to the question is esentially yes, the capabilities available to you will make a profound difference in what you can or can't do with your research.

Copyright 2008, Semantech Inc.

The importance of coaching to an organization

2008-03-18T12:22:00.000-07:00

I can’t speak for organizations I haven’t worked for, but I have never personally seen any extensive use of either coaching or mentoring in the workplace. I have had the opportunity to work as a contractor in about 20 organizations over the years. I’m relatively certain that some coaching / mentoring was occurring in some of those places, yet I never saw it in operation. I mention this as a disclaimer partly because most of the assignments I’ve had involve a certain expectation that resources when hired are ‘fully developed’ and simply don’t require any further HR investments.

On the other hand, there is a very real question as to whether all managers engage in or should engage in coaching and mentoring with their staff. As projects become more complex and staff more senior, there is an expectation that overt guidance is almost patronizing. Also, if one is dealing with contract or outsourced staff, what is the goal or value behind coaching (and mentoring makes no sense at that point)?

Where I see true coaching occurring is in the context of interactive oversight. This oversight is conducted by both mangers and senior IT staff. The goal in these cases is three-fold:

1 – To ensure team coordination and communication.

2 - To ensure project expectations are being followed through as tasks are being worked.

3 – In some cases to groom key personnel for taking leadership roles later in the same or other related projects.

I’m also not sure that there is really an accepted definition of what coaching is (as opposed to mentoring, which easier to define due to its one on one nature). I believe that certain aspects of project management overlap with what might be defined as coaching if an organization had a defined it formerly. One of the key differences might be the fact that your reporting chain was not involved in providing the actual coaching.

For the organizations I've worked with, both coaching and mentoring have been a very beneficial moves. The organizations have often been federal agencies and have suffered an employee exodus due to previous budget cuts and ongoing retirements. The number of remaining employees with vital system and process knowledge has now dwindled to the point that there are many areas where resources are only one deep. The only way I know of to facilitate rapid assimilation of complex knowledge is through close coordination and dialog with experts and that’s essentially what a coach or mentor is.

Usability Testing

2008-03-15T11:04:00.000-07:00

The development process for software or websites is often more complicated than we’d like to assume it is. The primary goal for usability testing is to ensure that the product being developed actually matches the intended audience’s expectations. Sometimes it is difficult to achieve this type of outcome without involving true end-users within the testing process somewhere. This also tends to imply that end-users were hopefully involved in other aspects of the development of the product as well. Accessibility testing is an extension to usability testing focused on more specific mechanical concerns.

Continuous Improvement

2008-03-15T10:57:00.000-07:00

What type of process is required to make sure that an organization is going through continuous improvement?

I don't think any process can ensure that an organization adopts or adheres to a Continuous Improvement (CI) paradigm. The issue is more related to the culture of the organization than any particular determination of what steps might be required to actualize CI. If a culture is not motivated to do it, or doesn't understand it or doesn't have an effective way to operationalize it - it fails. The more problematic consideration, though, is the fact that many organizations are specifically based upon sub-optimizing roles through specialization. In organizational cultures where the workforce traditionally has had little influence on their roles and have been trained for a very narrow range of specialization they become used to "staying within the bounds" laid forth by management. In such organizations standing out through extra levels of participation and innovation is often viewed as insubordination to the chain of command. Such cultures have little chance of deploying or sustaining CI paradigms.

There are differing opinions about what the "Problem" in problem solving actually means within the context of Total Quality Management (TQM). He notes that amongst most of the practitioners in the field it refers to defects and the ability to reduce or otherwise correct them. Problem-solving however is not a negative activity, in fact, problem-solving is something we all do in the course of our everyday lives constantly. Problems are the challenges by which we define ourselves - they are much more varied and complex than than those measured by the statistical methods we have available to us currently.

Problems in highly controlled environments / organizations are considered negative and are very narrowly defined and attacked. When these types of organizations face complex problems they attempt to break them down using the techniques they've become accustomed to and are overwhelmed. There is a balance that is needed for CI to take place, that balance includes a recognition that not all activity can be scientifically programmed, that our greatest strengths are in fact our ingenuity and flexibility.

I've witnessed a number of attempts to introduce CI by organizations that exhibit rigid cultures and none of them fared too well. Interestingly enough, the places I've seen that have working CI paradigms are the ones where there is still a great dependence on manual processes with significant quantities of experts in the loop.

Copyright 2008, Semantech Inc.

All about ECM

2008-03-15T10:20:00.000-07:00

Calling ECM designers and developers: What do Documentum, OpenText, Interwoven and other ECM technologies do better than Sharepoint and how do they maintain competitive advantage?

“In general, the COTS solutions you've listed have the following advantages over Sharepoint: 1 - Better scalability; i.e. designed somewhat to be extend across storage management solutions. 2 - More robust support for complex customizing development activities. 3 - They were around first; i.e. Sharepoint didn't become very user friendly until the 2003 release. 4 - The navigation paradigms are generally better in the other solutions. 5 - Sharepoint is very SQL Server focused and many large enterprises are Oracle shops... Having said all of that, I don't any of the major vendors have yet realized the true potential of ECM - it won't mature we have the ability to integrate analytics from common BI suites across unstructured data stores...”

Enterprise 2.0

2008-03-15T10:18:00.000-07:00

What is Enterprise 2.0 and does this really extends Web2.0 to enterprises?

As may be obvious, Enterprise 2.0 is a bit of a buzzword, however I'm not sure the definition is nailed down yet. My view of what it is or might be is much more aligned to the notion of unified management rather than exploitation of Web 2.0 technologies per se. Web 2.0 may certainly be a part of that but what is really key in determining any next generation enterprise solution is the ability to design and manage at the enterprise rather than the system, service or application level...”

Copyright 2008, Semantech Inc.

What are White Papers, Really?

2008-03-15T10:16:00.000-07:00

What was the original Purpose of an "IT White Paper" and are they an important resource document?

“A white paper is a bit of a hybrid between pure marketing and a formal research paper. They don't always turn out that way but that is the target domain. Pure marketing doesn't tend to provide users or potential clients with any meaningful information and research papers are difficult to read due to length, formatting and the insistence by academia that everyone needs to think like everyone else (i.e. the obsession w/ citations).

A white paper is an opportunity to present your visionary thoughts in your own context without just making a sales pitch. I've always felt the best sales pitch anyway is the ability to demonstrate your competence or expertise in the field.”

Computers, a Threat?

2008-03-15T10:14:00.000-07:00

Why do businesses consider computers a threat to their bottom line?

“You've captured some of the answer in your question... People generally feel somewhat unsecure or stressed with the notion of constant change. It is challenging for folks to continue to learn new skills every several years - the cycles for change have accelrated dramatically over the past 50 years in particular. I still remember my great grandmother explaining the workings of a hotel telephone switchboard from the 1920's - that technology took decades to change. I have to evaluate new products every other week. My grandmother never flew in a plane either - cultural biases are hard to bypass, sometimes they never are.”

In-Design & Web Development

2008-03-15T10:12:00.000-07:00

I have a client that has creates an Adobe InDesign publication each month that they wish to publish online using the export to DreamWeaver feature. Does anyone have experience with this solution or perhaps know of a better one?

“2 Considerations: 1 - If you're using the Adobe Master Collection, you can use the export feature (Indesign to Dreamweaver) once to establish the template and then publish new content later using Contribute (this won't allow for major template changes but is easy - quick). 2 - Publishing using using InDesign is better suited for creating PDFs - if the publication is meant o be a web publication it ought to be transferred over to another design tool (Dreamweaver etc.) The export to Dreamweaver will not be as effective as designing to the medium directly.”

Cost Effective ERP

2008-03-15T10:08:00.000-07:00

How can nonprofits integrate ERP in their operations? Is there any ERP software/service that is cost-effective and relevant for the nonprofit industry?

“The software can be significantly cheaper than that if you go open source: http://www.compiere.com/ . There are a number of other open source ERP products as well, the choice should largely depends upon the nature of what it is that the non-profit does as all are not alike.

The main cost will likely be for IT staff to configure it, and the risk and potential greatest cost would arise if the implementation wasn't done properly.

Copyright 2008, Semantech Inc.

Risk Tolerance & Recession

2008-03-15T10:06:00.000-07:00

Risk Tolerance in Tech Startups; what Impact will a Recession have on Attracting Top Talent?

“First of all, we're not officially in a recession yet - the fact that we haven't already plunged into one given a year of insane fuel costs and the melt-down of the mortage market is somewhat remarkable in itself. The other thing to consider is that "stability" as we have viewed it in the job market traditionally, simply no longer exists. There is an element of risk now in staying somewhere - one place - too long. If you aren't used to competing in the marketplace, having to learn it while in the process of being laid off is fairly daunting.

What guarantees are there in a recession that your ship will weather the storm or even it does will it do so with you or find cheaper alternatives overseas? There is also the reward side to consider - there will always be new technology focused markets emerging regardless of the economy for those with vision and the dedication to make that vision happen. The simple answer may be that the type of people who can make start -ups succeed will always be out there, it's what makes life exciting for them and perhaps new ventures may become increasingly attractive to those whose 'traditional' careers are becoming less 'assured' every day.”

Copyright 2008, Semantech Inc.

Question - Agile Security

2008-03-15T10:05:00.000-07:00

How is the Agile framework enforced and technologically secured?

“I'm not sure that the methodology employed (in any typical IT environment or Agile project) necessarily impacts the nature of a security solution per se. Some may posit that an Agile approach may cause certain vulnerabilities due to haste in design, development and deployment but this is not an absolute given...”

Welcome to Ensights

2008-03-15T10:01:00.000-07:00

Hello and welcome to the Ensights.

This Blog has been built as a corollary to our Ensights online magazine which is still under development. We will preview articles here on a wide variety of enterprise IT topics and also capture here portions of dialogs from other forums such as Linkedin.com. There is no single focus for this publication - which gives us the flexibility to muse on any issues great or small in our domain, we hope you enjoy it and thanks for stopping by...

Semantech Ensights

What should the top priorites be for our nation's first CTO?

Is there a Cultural Bias behind Innovation?

The Tools of Innovation part 1: Concept Maps

Metaphors for Design

Ideation

How do you choose a right architecture?

Links:

What does the term "Personalized Medicine" mean to you and in what way do you think it has the greatest potential?

Semantic use of H1 heading

IT Talent Shortage: Should IT executives do more to create local talent instead of just complaining about the trend?

What aspects are important to have in a "online newsroom" (social media newsroom)?

Can Artificial Intelligence support ad hoc decisions?

Will the Internet help bring about enlightened societies that actually think critically?

Is Radio & TV Advertising Dead?

Have you recently tried to hire a data quality team member?

Links:

What does Web 2.0 mean to you ?

What Business Intelligence trends have you observed?

Links:

How your firm match IT roles to ITIL roles?

What is your definition of a maturity model?

Using social networking tools for knowledge management

question:

Links:

Linkedin Question "SaaS vs. SOA"

How do you see Knowledge Management defining the current state of research in life sciences?

What project management tools do you like and why?

How to get start with SOA (Service Oriented Architecture)?

Will SaaS change the face of Enterprise Architecture and SOA ?

Team Performance

Project and Organizational ROI

Does Modeling have a place in Evolutionary Design?

Exploratory Testing & Agile Methodology

Business Analyst Salaries

Taxonomy Seminars

Knowledge Management & Life Sciences

The importance of coaching to an organization

Usability Testing

Continuous Improvement

All about ECM

Enterprise 2.0

What are White Papers, Really?

Computers, a Threat?

In-Design & Web Development

Cost Effective ERP

Risk Tolerance & Recession

Question - Agile Security

Welcome to Ensights