KM Programmes Captured by technology; Taxonomy (Retrieval rather than Search) #metadata #taxonomy #governance


ellwoods@btinternet.com <steve@...>
 

I'm currently kibitzing on the introduction of a new Knowledge/Collaboration strategy - my only involvement has been some input on the behaviours/culture side.

I note with some interest that the technical side - the "Architecture" team - are having separate meetings where they have decided what technical solution is going to be used; and on that basis have decided some key policies as to how the programme will behave. This has brought the team into open conflict with the governance team.

Is this an unusual experience?

As part of this, the technical team have "explained" that retrieval is better than search, so all Knowledge Assets will have to have metadata applied in accordance with the "approved taxonomy" before it can be filed. Leaving aside that where one individual expects to retrieve information from may not be where another thinks to store it - and the "approved taxonomy" may not fit...

What is the general experience of enforcing rigid metadata completion on the willing sharing of knowledge from teams?
Will the deprecation of search reduce the possibility of serendipitous discovery?

I'm surprised to discover that there is an approach of mandating asset ownership; good in theory, but when working with hundreds of thousands to millions of information items - which apparently are going to be manually migrated to the new system by the "the users" - of whom there also tens of thousands - a rigid system of enforcing ownership seems fraught with problems. Suggestions have been made to allocate to business units initially, but this seems unpopular.

How have others handled problems with document ownership in a rapid turnover organisation?


Tom Reamy <tomr@...>
 

First, if you have 100,000’s or millions  of information items and expect people to find what they are looking for without metadata is to condemn a lot of people to a really frustrating and wasteful experience.  Search without metadata of some sort is simply not even close to adequate so hats off to the technical group for demanding it.  I was just at a conference, Enterprise Search Summit, where the continued pain of badly implemented search (no metadata, no taxonomy) was extremely prevalent. 

 

However, there are a lot of problems associated with adding metadata and a taxonomy and you have pinpointed some of them, but some of your fears are fairly easily handled. For example, an official taxonomy that does not take individual variations into account.  There are a number of ways to deal with that from having rich thesauri behind the taxonomy to capture individual terminology variations to developing a faceted taxonomy to allow people to access information based on what they know, not some official one dimensional taxonomy. 

 

BTW, I wouldn’t worry about retrieval reducing the possibility of serendipitous discovery – a good search/retrieval interface should include at least one facet for this kind of discovery – usually driven by a clustering capability.

 

As far as the best way to get that much content tagged, I’ve seen quite a few approaches but the one that has worked more than any other is what is often called a hybrid approach.  You typically have 3 sources for tags – a central team of librarians and/or knowledge managers who know classification really well but not necessarily the variation of business terms, the users or subject matter experts who are the source for those business terms but hate having to add metadata, and lastly software that “automatically” tags documents but needs to have that capability built into it. 

 

So the hybrid answer I’ve seen work is having software that can suggest tags to humans – both users/SME’s and a library team.  The software rarely can do a good enough job without some human judgment, but having it suggest tags to users makes the process of tagging not only much better (users are often bad taggers) but much easier.  It is easier to get someone to simply say yes to a tag than think up one.  The library staff helps develop the categorization / tagging capability of the software although they often need help in getting started with the process (it’s a related but different skill for most), and they also can do some overall vetting of the tagging process as well as monitor and adjust the performance of the tags in the search/retrieval process.

 

That is a quick overview of the process that I’ve seen work better than any either/or approach (search or retrieval, manual or automated, users or librarians).

 

The model can be adapted to different situations but overall it is the best approach that I’ve seen.

 

Hope that helps,

 

Tom

 

 

 

 

 

Tom Reamy

Chief Knowledge Architect

KAPS Group, LLC

www.kapsgroup.com

510-530-8270 (O)

510-530-8272 (Fax)

510-333-2458 (M)

 

From: sikmleaders@... [mailto:sikmleaders@...] On Behalf Of ellwoods@...
Sent: Monday, May 31, 2010 7:02 AM
To: sikmleaders@...
Subject: [sikmleaders] KM Programmes Captured by technology; Taxonomy (Retrieval rather than Search)

 

 

I'm currently kibitzing on the introduction of a new Knowledge/Collaboration strategy - my only involvement has been some input on the behaviours/culture side.

I note with some interest that the technical side - the "Architecture" team - are having separate meetings where they have decided what technical solution is going to be used; and on that basis have decided some key policies as to how the programme will behave. This has brought the team into open conflict with the governance team.

Is this an unusual experience?

As part of this, the technical team have "explained" that retrieval is better than search, so all Knowledge Assets will have to have metadata applied in accordance with the "approved taxonomy" before it can be filed. Leaving aside that where one individual expects to retrieve information from may not be where another thinks to store it - and the "approved taxonomy" may not fit...

What is the general experience of enforcing rigid metadata completion on the willing sharing of knowledge from teams?
Will the deprecation of search reduce the possibility of serendipitous discovery?

I'm surprised to discover that there is an approach of mandating asset ownership; good in theory, but when working with hundreds of thousands to millions of information items - which apparently are going to be manually migrated to the new system by the "the users" - of whom there also tens of thousands - a rigid system of enforcing ownership seems fraught with problems. Suggestions have been made to allocate to business units initially, but this seems unpopular.

How have others handled problems with document ownership in a rapid turnover organisation?


Matt Moore <innotecture@...>
 

The term I have been pushing is "cybord metadata" (should have the first of a series of articles on that topic published in a couple of weeks). Basically my take is that we need a mix of experts, users & machines to manage our information environments - and there are distinct roles for each group. I don't believe that central taxonomies managed by humans are obsolete - but how they are constructed, applied and consumed is changing greatly. I suspect my position is pretty close to Tom's. Anyway, jetlag getting to me. Will post details of article when it's out - includes some nice examples from Hong Kong academics, Australia museums and OpenCalais.

BTW Get ready for a lot of hype about the semantic web (ontologies, RDF) imminently. People in PR and ad agencies have just discovered it and decided it is the next "social media". Batten down the metadata hatches!!!

Matt Moore
+61 423 784 504

On May 31, 2010, at 11:51 AM, "Tom Reamy" <tomr@...> wrote:

 

First, if you have 100,000’s or millions  of information items and expect people to find what they are looking for without metadata is to condemn a lot of people to a really frustrating and wasteful experience.  Search without metadata of some sort is simply not even close to adequate so hats off to the technical group for demanding it.  I was just at a conference, Enterprise Search Summit, where the continued pain of badly implemented search (no metadata, no taxonomy) was extremely prevalent. 

 

However, there are a lot of problems associated with adding metadata and a taxonomy and you have pinpointed some of them, but some of your fears are fairly easily handled. For example, an official taxonomy that does not take individual variations into account.  There are a number of ways to deal with that from having rich thesauri behind the taxonomy to capture individual terminology variations to developing a faceted taxonomy to allow people to access information based on what they know, not some official one dimensional taxonomy. 

 

BTW, I wouldn’t worry about retrieval reducing the possibility of serendipitous discovery – a good search/retrieval interface should include at least one facet for this kind of discovery – usually driven by a clustering capability.

 

As far as the best way to get that much content tagged, I’ve seen quite a few approaches but the one that has worked more than any other is what is often called a hybrid approach.  You typically have 3 sources for tags – a central team of librarians and/or knowledge managers who know classification really well but not necessarily the variation of business terms, the users or subject matter experts who are the source for those business terms but hate having to add metadata, and lastly software that “automatically” tags documents but needs to have that capability built into it. 

 

So the hybrid answer I’ve seen work is having software that can suggest tags to humans – both users/SME’s and a library team.  The software rarely can do a good enough job without some human judgment, but having it suggest tags to users makes the process of tagging not only much better (users are often bad taggers) but much easier.  It is easier to get someone to simply say yes to a tag than think up one.  The library staff helps develop the categorization / tagging capability of the software although they often need help in getting started with the process (it’s a related but different skill for most), and they also can do some overall vetting of the tagging process as well as monitor and adjust the performance of the tags in the search/retrieval process.

 

That is a quick overview of the process that I’ve seen work better than any either/or approach (search or retrieval, manual or automated, users or librarians).

 

The model can be adapted to different situations but overall it is the best approach that I’ve seen.

 

Hope that helps,

 

Tom

 

 

 

 

 

Tom Reamy

Chief Knowledge Architect

KAPS Group, LLC

www.kapsgroup.com

510-530-8270 (O)

510-530-8272 (Fax)

510-333-2458 (M)

 

From: sikmleaders@yahoogroups.com [mailto:sikmleaders@yahoogroups.com] On Behalf Of ellwoods@btinternet.com
Sent: Monday, May 31, 2010 7:02 AM
To: sikmleaders@yahoogroups.com
Subject: [sikmleaders] KM Programmes Captured by technology; Taxonomy (Retrieval rather than Search)

 

 

I'm currently kibitzing on the introduction of a new Knowledge/Collaboration strategy - my only involvement has been some input on the behaviours/culture side.

I note with some interest that the technical side - the "Architecture" team - are having separate meetings where they have decided what technical solution is going to be used; and on that basis have decided some key policies as to how the programme will behave. This has brought the team into open conflict with the governance team.

Is this an unusual experience?

As part of this, the technical team have "explained" that retrieval is better than search, so all Knowledge Assets will have to have metadata applied in accordance with the "approved taxonomy" before it can be filed. Leaving aside that where one individual expects to retrieve information from may not be where another thinks to store it - and the "approved taxonomy" may not fit...

What is the general experience of enforcing rigid metadata completion on the willing sharing of knowledge from teams?
Will the deprecation of search reduce the possibility of serendipitous discovery?

I'm surprised to discover that there is an approach of mandating asset ownership; good in theory, but when working with hundreds of thousands to millions of information items - which apparently are going to be manually migrated to the new system by the "the users" - of whom there also tens of thousands - a rigid system of enforcing ownership seems fraught with problems. Suggestions have been made to allocate to business units initially, but this seems unpopular.

How have others handled problems with document ownership in a rapid turnover organisation?


 


Stephanie Barnes
 

Hi,

I wanted to comment on the division between IT and everyone else that you mentioned. No, it is not unusual, unfortunately; however, it is a recipe for disaster--frustration, wasted time, and money and even more animosity between Business and IT.

Business and IT have to work together to find a solution that meets everyone's needs, that's the only way it will work. It takes a lot of work/facilitation/communication, but is worth it in the end to have a solution that the users will use and IT will support.

Good luck,
Stephanie

Stephanie Barnes, BBA, MBA
Chief Chaos Organizer
Missing Puzzle Piece Consulting
www.missingpuzzlepiececonsulting.ca

Authorised licensee of Knoco Ltd. www.knoco.com


Tom <tman9999@...>
 

--- In sikmleaders@yahoogroups.com, "ellwoods@..." <steve@...> wrote:

I note with some interest that the technical side - the "Architecture" team - are having separate meetings where they have decided what technical solution is going to be used; and on that basis have decided some key policies as to how the programme will behave. This has brought the team into open conflict with the governance team.

Is this an unusual experience?
I have to laugh, Steve, because this is so typical it is almost comical.

Biz Mgr: "We need to something about knowledge management."
IT Dir: 'ok - good, I'll jump right on it.'
BM: we have too many documents and files, and we can never find anything.
ITD: hmmmm...sounds like we should probably put in Sharepoint. I'll need $250k for licences and the project allowance to get v1 up.
BM: what's Sharepoint?
....

Or something like that.

Many organizations I've worked with or heard about who try to do something on the KM front initially focus on IT-based, and therefore owned and led, solution approaches.

Unfortunately as we all now know, most of KM is at root behavioral based and process-driven. The elegantly defined solution diagrams would function beautifully if it weren't for those pesky employees who use the system in ways it was never intended, and can't seem to follow simple procedures (like adding the correct tags to the documents they wish to add).

One way to think about the challenge you describe is to reduce it down to an execution strategy problem. Using this approach the governance team needs to allocate decision rights in a way that will ensure the proposed solution is implemented appropriately and, more important, adopted by the target audience in a way that addresses the challenge that prompted the thing in the first place. If the governance team is IT led or dominated, then there will be a natural tendency to elevate IT's role. The business should have the dominant say. But to do that effectively the business leaders need to have some level of sophistication about what KM is, and how it can be used to address the business issue they have decided to target. Chicken and egg, almost.

So what can you do? Educate the business leadership about what real KM is about and how successful solutions are developed, implemented, and measured. Educate everyone about the importance of decision rights, and ensure that these are allocated in a way that best serves the business, not IT's self-interest.

I'm sure others have thoughts and experiences that they can share that would add to this list. The good news is it sounds like you are at an early enough stage to influence the direction of this effort for the better. Good luck!! Let us know how it goes.

-Tom

Tom Short Consulting
Knowledge Transfer
Process Analysis
Metrics


Fred Nickols
 

I've followed this short thread and thought I'd pick up behind Stephanie's post...

The tension and contention between IT and users is hardly new; it goes on all the time. However, when the head-butting gets really severe, someone higher up usually steps in to resolve the conflict. This case makes wonder if there is no real sponsor or owner for the KM initiative higher in the hierarchy than the conflicted parties. If there is no higher sponsor/owner, then Stephanie is dead on: "it is a recipe for disaster--frustration, wasted time, and money and even more animosity between Business and IT."

Fred Nickols
fred@nickols.us

--- In sikmleaders@yahoogroups.com, "Barnes" <stephanie.barnes@...> wrote:



Hi,

I wanted to comment on the division between IT and everyone else that you mentioned. No, it is not unusual, unfortunately; however, it is a recipe for disaster--frustration, wasted time, and money and even more animosity between Business and IT.

Business and IT have to work together to find a solution that meets everyone's needs, that's the only way it will work. It takes a lot of work/facilitation/communication, but is worth it in the end to have a solution that the users will use and IT will support.

Good luck,
Stephanie

Stephanie Barnes, BBA, MBA
Chief Chaos Organizer
Missing Puzzle Piece Consulting
www.missingpuzzlepiececonsulting.ca

Authorised licensee of Knoco Ltd. www.knoco.com


Simard, Albert <albert.simard@...>
 

There was no signature on the posting, and I can’t simply hit reply, so to whom it may concern:

 

The bottom line is that if people are forced to use IM processes that don’t fit their ways of working, they will find ways to work around or bypass whatever system you implement.  And if it takes more than a minute (30 seconds is better) to file something, they won’t.  Period.  And the harder you try to force them to do what they don’t want to do, the more energy they’ll expend figuring out how not to do it.  The expected outcome of continuing along the path that you describe is yet another system that failed.

 

This gets to what I used to see as two sides of a coin, but has now evolved to a 3-legged stool.  First, every organization has two dimensions – running itself and serving clients.  Without an organization, there is no way to serve clients and without clients an organization has no reason to exist.  The problem is that Information architectures designed for running an organization are of little value for serving clients and vice versa.  One size cannot fit all.  And neither side of the organization understands the other.  Enter social networking, tag clouds, wisdom of crowds and unstructured taxonomies.  Fun to browse and be creative, but there is no way that the searcher can be assured that they have found everything that they’re looking for.

 

The end result is that two or three types of structures have to be managed (Yikes!).  Running the organization requires a fairly rigid, authoritative taxonomic structure.  Everything has to be found, it must be consistent, it must be aggregateable to higher levels, and it must follow enterprise reporting structures.  Most users care little about all that.  Information must also be structured in the way people use it.  In science, this often involves international global standard taxonomies (not someone’s favourite terms, as IT often misbelieves).  Thus, we need a user-oriented taxonomy.  And those who actually use systems know this far better than any IT group I’ve ever encountered. 

 

Finally, there should be a way to do unstructured searches, to see what may be in there that doesn’t fit the official taxonomies.  This will support the tacit (creative) side of the house.   This will also enable people to find emerging content that hasn’t been officially classified yed.

 

Albert J. Simard, Ph.D.

Knowledge Manager / Gestionnaire du savoir

 

Defence R&D Canada - / R&D pour la defense Canada

305 Rideau St., 9th floor - AH11 / 305 rue Rideau, 9 ieme etage -AH11

Ottawa, Ontatio K1A 0K2

Canada

Tel: 613-943-3501   Fax: 613-996-7063

e-mail: albert.simard@...

 


Marie-Michelle Strah <marie.strah@...>
 

First time posting here - have been lurking and thoroughly enjoying (gathering knowledge about group dynamics and folksonomies!)

Agree this is business/IT gap issue - but I've seen the other side of the coin, unfortunately. KM or functional leads embark on knowledge management initiatives without ever consulting IT - even to the point of choosing tools to enable knowledge and data sharing. The result in that case is disastrous - since often the platforms chosen (in example below it is SharePoint, but fill in the blank) is neither appropriate or has the featues and/or functionality desired. Lots of time and money wasted on "square peg in round hole" efforts that give a bad name both to KM and the platform - and give the IT folks a big headache.

I work both the KM/governance and technical aspects of our projects to ensure all aspects adequately addressed at project inception. However, am currently involved in an untangling effort where a tool that uses outstanding data visualization (Silverlight) was chosen for its "pizzazz" by non-technical functional experts but for many reasons (data architecture being the primary) is ill-suited for the project (it was originally designed to manage commercial facilities data and has a hard time with taxonomies and classifications in healthcare context).

So... it goes both ways.

Thanks for great comments and convo on this!

Best,
Michelle

Marie-Michelle Strah, PhD
Manager, Information Systems
General Dynamics Information Technology
marie.strah@...
michelle.strah@...



> I note with some interest that the technical side - the "Architecture" team - are having separate meetings where they have decided what technical solution is going to be used; and on that basis have decided some key policies as to how the programme will behave. This has brought the team into open conflict with the governance team.
>
> Is this an unusual experience?

Biz Mgr: "We need to something about knowledge management."
IT Dir: 'ok - good, I'll jump right on it.'
BM: we have too many documents and files, and we can never find anything.
ITD: hmmmm...sounds like we should probably put in Sharepoint. I'll need $250k for licences and the project allowance to get v1 up.
BM: what's Sharepoint?
....

Or something like that.

_._,_.___
Recent Activity:
.



ellwoods@btinternet.com <steve@...>
 

Just a brief follow up to say thanks - and add a signature!

I'm relieved - and to be honest - unsurprised to find that the tension between the IT & governance teams is not that unusual.

I'm all over the fact that some taxonomy needs to be there; I think it might be a bit tardy to be thinking about that when we've apparently decided what our solution is!

I'll let you know how it goes - and if I manage to influence our direction.


Steve Ellwood
http://steveellwood.com
http://twitter.com/steveellwood [occasional, peripatetic @KMers]
Knowledge Management, Solution & Service Design

--- In sikmleaders@yahoogroups.com, "ellwoods@..." <steve@...> wrote:

I'm currently kibitzing on the introduction of a new Knowledge/Collaboration strategy - my only involvement has been some input on the behaviours/culture side.

I note with some interest that the technical side - the "Architecture" team - are having separate meetings where they have decided what technical solution is going to be used; and on that basis have decided some key policies as to how the programme will behave. This has brought the team into open conflict with the governance team.

Is this an unusual experience?

As part of this, the technical team have "explained" that retrieval is better than search, so all Knowledge Assets will have to have metadata applied in accordance with the "approved taxonomy" before it can be filed. Leaving aside that where one individual expects to retrieve information from may not be where another thinks to store it - and the "approved taxonomy" may not fit...

What is the general experience of enforcing rigid metadata completion on the willing sharing of knowledge from teams?
Will the deprecation of search reduce the possibility of serendipitous discovery?

I'm surprised to discover that there is an approach of mandating asset ownership; good in theory, but when working with hundreds of thousands to millions of information items - which apparently are going to be manually migrated to the new system by the "the users" - of whom there also tens of thousands - a rigid system of enforcing ownership seems fraught with problems. Suggestions have been made to allocate to business units initially, but this seems unpopular.

How have others handled problems with document ownership in a rapid turnover organisation?


Douglas Weidner
 

Michele,

 

Insightful comments.

 

I myself have seen enthusiastic KMers, past students, who have much KM passion and some tech savvy, but unfortunately not enough to consider scalability, compatibility, etc. – bunch of other strategic –ibilities.

 

We now address that issue in our KM Certification Program as one of many key concerns.

 

Douglas Weidner, eCKM Mentor

Chairman, International Knowledge Management Institute

Best in KM Training & Certification

Home of the KM Body of Knowledge (KMBOK)™,and

the Knowledge Maturity Model (KMM)™

www.kminstitute.org

703-757-1395

 

 

 

From: sikmleaders@... [mailto:sikmleaders@...] On Behalf Of Marie-Michelle Strah
Sent: Wednesday, June 02, 2010 12:12 AM
To: sikmleaders@...
Subject: Re: [sikmleaders] Re: KM Programmes Captured by technology; Taxonomy (Retrieval rather than Search)

 

 

First time posting here - have been lurking and thoroughly enjoying (gathering knowledge about group dynamics and folksonomies!)

Agree this is business/IT gap issue - but I've seen the other side of the coin, unfortunately. KM or functional leads embark on knowledge management initiatives without ever consulting IT - even to the point of choosing tools to enable knowledge and data sharing. The result in that case is disastrous - since often the platforms chosen (in example below it is SharePoint, but fill in the blank) is neither appropriate or has the featues and/or functionality desired. Lots of time and money wasted on "square peg in round hole" efforts that give a bad name both to KM and the platform - and give the IT folks a big headache.

I work both the KM/governance and technical aspects of our projects to ensure all aspects adequately addressed at project inception. However, am currently involved in an untangling effort where a tool that uses outstanding data visualization (Silverlight) was chosen for its "pizzazz" by non-technical functional experts but for many reasons (data architecture being the primary) is ill-suited for the project (it was originally designed to manage commercial facilities data and has a hard time with taxonomies and classifications in healthcare context).

So... it goes both ways.

Thanks for great comments and convo on this!

Best,
Michelle

Marie-Michelle Strah, PhD
Manager, Information Systems
General Dynamics Information Technology
marie.strah@...
michelle.strah@...

 


 

> I note with some interest that the technical side - the "Architecture" team - are having separate meetings where they have decided what technical solution is going to be used; and on that basis have decided some key policies as to how the programme will behave. This has brought the team into open conflict with the governance team.
>
> Is this an unusual experience?

Biz Mgr: "We need to something about knowledge management."
IT Dir: 'ok - good, I'll jump right on it.'
BM: we have too many documents and files, and we can never find anything.
ITD: hmmmm...sounds like we should probably put in Sharepoint. I'll need $250k for licences and the project allowance to get v1 up.
BM: what's Sharepoint?
....

Or something like that.

_._,_.___

Recent Activity:

.

Error! Filename not specified.

 


Steve Ardire <sardire@...>
 

KM or functional leads embark on knowledge management initiatives without ever consulting IT - even to the point of choosing tools to enable knowledge and data sharing.

Yup and would also add usually without ever consulting with IA / EIM people. I crossed this bridge some time ago and now suggest check out MIKE2.0 (Method for an Integrated Knowledge Environment)  an Open Source methodology for Enterprise Information Management that provides a framework for Information Development. 


Steve Ardire
sardire@...
360-868-4435 ( Google Voice # )
http://www.linkedin.com/in/sardire
skype: sardire / twitter: @sardire


On Tue, Jun 1, 2010 at 9:12 PM, Marie-Michelle Strah <marie.strah@...> wrote:
KM or functional leads embark on knowledge management initiatives without ever consulting IT - even to the point of choosing tools to enable knowledge and data sharing.


Marie-Michelle Strah <marie.strah@...>
 

Steve said:



Steve -

Thank you for the referral to the MIKE2.0 open standards - this will be extremely helpful in tackling our issues and integrates very nicely with agile development methodologies.

Douglas - info on new requirements for KM certification very heartening - will look into that.

Great discussion - nice to see others crossing IT/business chasm.

Best,
Michelle

Marie-Michelle Strah, PhD
General Dynamics Information Technology   
marie.strah@...
michelle.strah@...





From: Steve Ardire
To: sikmleaders@...
Sent: Wed, June 2, 2010 10:12:32 AM
Subject: Re: [sikmleaders] Re: KM Programmes Captured by technology; Taxonomy (Retrieval rather than Search)

 

KM or functional leads embark on knowledge management initiatives without ever consulting IT - even to the point of choosing tools to enable knowledge and data sharing.


Yup and would also add usually without ever consulting with IA / EIM people. I crossed this bridge some time ago and now suggest check out MIKE2.0 (Method for an Integrated Knowledge Environment)  an Open Source methodology for Enterprise Information Management that provides a framework for Information Development. 


Steve Ardire
sardire@gmail. com
360-868-4435 ( Google Voice # )
http://www.linkedin.com/in/sardire
skype: sardire / twitter: @sardire



Steve Ardire <sardire@...>
 

Hi Michele - you're most welcome and also see http://www.mkbergman.com/881/two-presentations-at-semtech-2010/ 

On Wed, Jun 2, 2010 at 6:26 PM, Marie-Michelle Strah <marie.strah@...> wrote:
 

Steve -

Thank you for the referral to the MIKE2.0 open standards - this will be extremely helpful in tackling our issues and integrates very nicely with agile development methodologies.

Douglas - info on new requirements for KM certification very heartening - will look into that.

Great discussion - nice to see others crossing IT/business chasm.

Best,
Michelle

Marie-Michelle Strah, PhD
General Dynamics Information Technology   
marie.strah@...
michelle.strah@...





From: Steve Ardire <sardire@...>
To: sikmleaders@...
Sent: Wed, June 2, 2010 10:12:32 AM

Subject: Re: [sikmleaders] Re: KM Programmes Captured by technology; Taxonomy (Retrieval rather than Search)

 

KM or functional leads embark on knowledge management initiatives without ever consulting IT - even to the point of choosing tools to enable knowledge and data sharing.


Yup and would also add usually without ever consulting with IA / EIM people. I crossed this bridge some time ago and now suggest check out MIKE2.0 (Method for an Integrated Knowledge Environment)  an Open Source methodology for Enterprise Information Management that provides a framework for Information Development. 


Steve Ardire
sardire@gmail. com
360-868-4435 ( Google Voice # )
http://www.linkedin.com/in/sardire
skype: sardire / twitter: @sardire




Michael Dieterle
 

Albert,

You make very good points about the challenge of serving two distinct dimensions and the need for a user-oriented taxonomy. But I can't quite agree with the statement that it cannot "be assured that they have found everything that they're looking for" when one searches based on folksonomies/tags. It might not be as reliable as a search for traditional content following the corporate taxonomy. But because you cannot predict the intent of the searcher, you won't know if one or the other result sets are more relevant. Especially with the increase in user generated content on internal portals, the likelihood of finding relevant search results in more current and user-tagged content will rise sharply.

Any other opinions on this recurring controversy?

Best Regards,
Michael Dieterle

VP of Knowledge Management, arcplan


Steve Ellwood <steve@...>
 

Michael,

I'm really enjoying the responses my question got, and the direction we're moving in.

You've said "Especially with the increase in user generated content on internal portals, the likelihood of finding relevant search results in more current and user-tagged content will rise sharply".

What's the expected growth in tagging compared to the growth in user generated content - my understanding was that there was usually a significant lag in getting growth in tagging - and therefore a lag in effective search using folksonomy.

Does this take us back towards a cultural shift again?


Steve Ellwood

http://steveellwood.com
http://twitter.com/steveellwood [occasional, peripatetic @KMers]
Knowledge Management, Solution & Service Design



On 4 June 2010 05:16, dieterle_michael <mvdieterle@...> wrote:
 

Albert,


You make very good points about the challenge of serving two distinct dimensions and the need for a user-oriented taxonomy. But I can't quite agree with the statement that it cannot "be assured that they have found everything that they're looking for" when one searches based on folksonomies/tags. It might not be as reliable as a search for traditional content following the corporate taxonomy. But because you cannot predict the intent of the searcher, you won't know if one or the other result sets are more relevant. Especially with the increase in user generated content on internal portals, the likelihood of finding relevant search results in more current and user-tagged content will rise sharply.

Any other opinions on this recurring controversy?

Best Regards,
Michael Dieterle

VP of Knowledge Management, arcplan



Simard, Albert <albert.simard@...>
 

Michael –

 

Thanks for your comments.  It’s all about the purpose of a search.

 

By “finding everything available,” I was discussing authoritative hierarchical functions, such as running an organization.  If I want to report on project or program progress or accomplishments at an organizational scale, I have to be able to find content on every project and it has to use standard metadata in a standard template so that I can combine individual reports into program and organizational totals.  This is bordering on records management.

 

If, on the other hand, I am doing a literature review for a paper, I have had great success with unstructured searches.  I use these in addition to structured library searches because they include cutting-edge stuff that hasn’t been published yet.  I find plenty of material and it often takes me in unanticipated directions.  Here, it is not necessary to find everything, but only to be confident that I’ve found the important material.

 

Al Simard


Simard, Albert <albert.simard@...>
 

Steve –

 

There’s another important dimension to structured vs unstructured searches – quality assurance and quality control.

 

For example, I work for Defence Research & Development Canada.  When we apply knowledge, people’s lives may be at stake.  As the battery commercial says “It simply has to work.”  We have to be very careful that the knowledge has been thoroughly validated.  One way to increase confidence is through scientific publication.  This provides “scientific authority” through peer review and journal editing.  This is a good place from which to begin real-world evaluation. 

 

When I post material on the Web, it’s a personal opinion, not authoritative.  Just because someone writes something doesn’t make it right.  Although I know enough to evaluate the quality of what I read, not all practitioners do.  For example, I’ve posted enough on Wikipedia to have encountered many of their QA/QC roving bots and participant comments to have reasonable faith in content that survives the process!  However, I also hear that there’s a lot of questionable material (generally in areas that I don’t tend to frequent).  So, although I often search the Web as a source of ideas and informed opinions, I don’t confuse what I find with authoritative knowledge.

 

Al Simard


Keith De La Rue
 

Steve -

The content part of my experience in KM may not be totally relevant to your work, but here are some of my thoughts just in case they are helpful.

Among other parts of my KM role over 8 years in a large organisation in Australia, I looked after an online knowledge library.  The library held about 3,500 data items, 95% of which included a single attached file.  These were mostly in Word or PowerPoint form, but could also be any type of file, including audio and video.  The content was product information, and the audience for the content was a B2B sales force.

Addressing your points here:

 - It seems rather strange that you would have different teams working on this - all of our administration of content was done by the one team.

 - Content was loaded by contributing Subject Matter Experts.  There were around 250 of these.  In order to load any content, the contributor had to also fill in the metadata fields.  These were pre-defined and fairly minimal; our contributors had no problem doing this.  Some of the metadata was automated, based on the fields filled in by the contributors.

 - Retrieval was both on by using the fairly minimalist taxonomy, as well as by search.  The search allowed a few different levels of refinement, either by using some of the metadata as a filter, or by full text.

 - Every document was always "owned" by a single contributor, who was accountable for keeping it up to date.  Each contributor would be reminded to do this every 3 months.  We always saw this as critical, and ensured that it was carried out, using a number of layers of both automated and manual process.  Content that was not updated was removed from audience access.

If this is of interest, you can refer to the slide pack I presented at the SI-KM April phone conference, or to my web site.  The slide pack is here:

http://www.slideshare.net/kdelarue/knowledge-transfer-toolkit-program

There is also a summary in this document:

http://delarue.net/blog/wp-content/uploads/2009/07/foundations-of-knowledge-idm-may-jun-09.pdf

Regards,
 
 - Keith.
--------------------------------------------------------
Keith De La Rue
AcKnowledge Consulting
...acting on knowledge, communication and learning
email:
keith@...
phone: +61 418 51 7676

blog: http://acknowledgeconsulting.com/
--------------------------------------------------------