Topics

Identify subject matter experts by information consumption? #expertise-location


Sam Yip
 

Hi all - I am curious to understand the process by which you would identify a subject matter expert within your organization. I am working on a computer science research on identifying experts by reference to their information consumption pattern -- essentially looking at the topics of what one would read/write, and use that as a proxy to determine if he is an expert**. For example, if someone keeps reading and writing about "5G automation" and "carrier aggregation" (from his emails, blogs, documents, presentations etc.) then he is potentially an expert on these topics. Is this a simplistic way to approach a nuanced task? Do you have other processes to determine if someone is an expert? 

Looking forward to your thoughts

**the research methodology here is to apply algorithms to retrieve information from different channels (along with who writes/reads what), and detect pre-defined topics from the body of information.


David Graffagna
 

Hi Sam ... a while ago I asked a question about establishing an SME Network and got some great input from the group. Some of this exchange may be of help. https://sikm.groups.io/g/main/message/6673


Dan Ranta
 

Hi Sam - the first point I want to share with you on this is that I have found it's important to use words carefully.  By that I mean that the word "expert" can be a tricky one.  I like to generally refer to employees having "expertise."  In many organizations, there is a special carve out category for an expert that has to do with official sanctioning from talent management / HR competency (really a process often between senior management and HR).  At GE, for example we had a very special category of expertise called Control Title Holders or CTHs.  I always considered these folks to be "experts" and it created a nice scenario where we could create a distinction between expert and expertise.  In general, it's very healthy for a KM program to define expertise and you also want to do so without creating any animosity between colleagues.  It's far easier to get massive uptake when you talk in terms of "expertise" and it's massive uptake and participation that you will want.  In summary, the word expertise is far softer and easier to promote.  Some other brief thoughts that deserve further expansion:

- Taxonomy is important to create topics and sub-topics for folks to select expertise
- Making your taxonomy look more and more like capabilities and competencies over time is challenging but essential to take it to the next level(s)
- KM is largely about processes (flows of knowledge) and having a large body of defined expertise is key to making knowledge flows of all types more precise and personalized...get the best answers...avoid collaborative overload...and much more
- Lastly, the only way to get to a large amount of expertise defined is to trust employees to define their own expertise; count on emotional intelligence to ensure completeness and accuracy of defining expertise

There is so much more - but I hope this helps.

Dan 

On Tue, Feb 23, 2021 at 8:11 AM Sam Yip <sam@...> wrote:
Hi all - I am curious to understand the process by which you would identify a subject matter expert within your organization. I am working on a computer science research on identifying experts by reference to their information consumption pattern -- essentially looking at the topics of what one would read/write, and use that as a proxy to determine if he is an expert**. For example, if someone keeps reading and writing about "5G automation" and "carrier aggregation" (from his emails, blogs, documents, presentations etc.) then he is potentially an expert on these topics. Is this a simplistic way to approach a nuanced task? Do you have other processes to determine if someone is an expert? 

Looking forward to your thoughts

**the research methodology here is to apply algorithms to retrieve information from different channels (along with who writes/reads what), and detect pre-defined topics from the body of information.



--
Daniel Ranta
Mobile:  603 384 3308


Simon Denton
 

That's certainly an approach we use to identify potential SME's. We combine that with self assessment, peer reviews and actions by others.

For example, our graduate engineers are largely responsible for content creation but the SMEs are responsible for checking and approval of the content. An SME might not use a Topic with sufficient frequency to be detected but the value they bring as mentors, checkers and approvers is a key indicator.

Regards,

Simon


From: main@SIKM.groups.io <main@SIKM.groups.io> on behalf of Dan Ranta via groups.io <danieleranta@...>
Sent: Tuesday, February 23, 2021 2:51:55 PM
To: main@sikm.groups.io <main@sikm.groups.io>
Subject: Re: [SIKM] Identify subject matter experts by information consumption?
 
Hi Sam - the first point I want to share with you on this is that I have found it's important to use words carefully.  By that I mean that the word "expert" can be a tricky one.  I like to generally refer to employees having "expertise."  In many organizations, there is a special carve out category for an expert that has to do with official sanctioning from talent management / HR competency (really a process often between senior management and HR).  At GE, for example we had a very special category of expertise called Control Title Holders or CTHs.  I always considered these folks to be "experts" and it created a nice scenario where we could create a distinction between expert and expertise.  In general, it's very healthy for a KM program to define expertise and you also want to do so without creating any animosity between colleagues.  It's far easier to get massive uptake when you talk in terms of "expertise" and it's massive uptake and participation that you will want.  In summary, the word expertise is far softer and easier to promote.  Some other brief thoughts that deserve further expansion:

- Taxonomy is important to create topics and sub-topics for folks to select expertise
- Making your taxonomy look more and more like capabilities and competencies over time is challenging but essential to take it to the next level(s)
- KM is largely about processes (flows of knowledge) and having a large body of defined expertise is key to making knowledge flows of all types more precise and personalized...get the best answers...avoid collaborative overload...and much more
- Lastly, the only way to get to a large amount of expertise defined is to trust employees to define their own expertise; count on emotional intelligence to ensure completeness and accuracy of defining expertise

There is so much more - but I hope this helps.

Dan 

On Tue, Feb 23, 2021 at 8:11 AM Sam Yip <sam@...> wrote:
Hi all - I am curious to understand the process by which you would identify a subject matter expert within your organization. I am working on a computer science research on identifying experts by reference to their information consumption pattern -- essentially looking at the topics of what one would read/write, and use that as a proxy to determine if he is an expert**. For example, if someone keeps reading and writing about "5G automation" and "carrier aggregation" (from his emails, blogs, documents, presentations etc.) then he is potentially an expert on these topics. Is this a simplistic way to approach a nuanced task? Do you have other processes to determine if someone is an expert? 

Looking forward to your thoughts

**the research methodology here is to apply algorithms to retrieve information from different channels (along with who writes/reads what), and detect pre-defined topics from the body of information.



--
Daniel Ranta
Mobile:  603 384 3308


Stephen Bounds
 

Hi Sam,

I would recommend also looking into social network analysis, and being cognisant of the distinction between an "expert" and a "knowledge broker".

Someone may be a trusted referrer without necessarily being an expert, and vice versa. Both can be very useful but I think are qualitatively different concepts.

Cheers,
Stephen.

====================================
Stephen Bounds
Executive, Information Management
Cordelta
E: stephen.bounds@...
M: 0401 829 096
====================================
On 24/02/2021 12:11 am, Sam Yip wrote:

Hi all - I am curious to understand the process by which you would identify a subject matter expert within your organization. I am working on a computer science research on identifying experts by reference to their information consumption pattern -- essentially looking at the topics of what one would read/write, and use that as a proxy to determine if he is an expert**. For example, if someone keeps reading and writing about "5G automation" and "carrier aggregation" (from his emails, blogs, documents, presentations etc.) then he is potentially an expert on these topics. Is this a simplistic way to approach a nuanced task? Do you have other processes to determine if someone is an expert? 

Looking forward to your thoughts

**the research methodology here is to apply algorithms to retrieve information from different channels (along with who writes/reads what), and detect pre-defined topics from the body of information.


Jonathan Gordon-Till
 

Hi Sam

First, just to reiterate Dan's comment about 'expert' versus 'expertise'. In my organization we take this a step further and consider 'experience' as part of the same spectrum, on the basis that even a complete novice has valuable experience from which others can learn.

Regarding your key question: Not exactly the answer you wanted, but there is AI technology such as Starmind (starmind.ai) which develops a neural network of connections between 'concepts' and 'people' based on the digital content created by the people. So it's not looking at what documents they are reading, but only what they are creating - and making a broad assumption that content creation somehow equates to experience. The more a person creates content relating to a concept, the more he/she is associated with that concept, hence the greater the indicator of experience (which may be roughly equated with expertise, etc.). So in principle, you can use e.g. Starmind technology to 'find an expert'.

Regards

Jonathan / UK


Sam Yip
 

Thank you all for the great comments and insights. A lot of valuable insights and there are a few points I’d love to learn more about:  

 

  • How do you determine the granularity of the taxonomy, which will be used as topics? In general is it always better to have fine-grained terms for taxonomy e.g. “SharePoint data migration” (fine-grained) vs “Information management” (very broad).  Also how do you strike the balance of letting taxonomy evolve on its own (which may be pulled in different directions by different offices across different regions) vs having a closely monitored and centralised taxonomy (at the risk of not keeping up with evolution of expertise/knowledge)?
  • it makes a lot of sense to differentiate experts - expertise - experience. How do you capture experience, or even just familiarity of a process, short of having a full-blown CoP or Q&A forum which requires a coordinated effort? To illustrate by way of example, say I am a project manager at a healthcare corporation at its Chicago office, and have previously looked into the process of applying for patents in Norway. I have spent 40 hours gathering information & requirements before the project was put on hold, but my effort will still be useful for for someone in say London office who is looking to apply for patents in Norway. How do you transfer that? 

  • In your SME networks, is there usually an inherent bias for senior staff, because they have built up sufficient expertise on topics? 


Murray Jennex
 

we used to call what Stamind does social network analysis and that is a powerful way of seeing who people email or possibly call when they have questions.  Of course this doesn't work for environments where people can just walk to the person they think knows the answer.  SNA is difficult to do and keep current and that is the only drawback....murray jennex


-----Original Message-----
From: Jonathan Gordon-Till <jonathan@...>
To: main@SIKM.groups.io
Sent: Tue, Feb 23, 2021 3:01 pm
Subject: Re: [SIKM] Identify subject matter experts by information consumption?

Hi Sam

First, just to reiterate Dan's comment about 'expert' versus 'expertise'. In my organization we take this a step further and consider 'experience' as part of the same spectrum, on the basis that even a complete novice has valuable experience from which others can learn.

Regarding your key question: Not exactly the answer you wanted, but there is AI technology such as Starmind (starmind.ai) which develops a neural network of connections between 'concepts' and 'people' based on the digital content created by the people. So it's not looking at what documents they are reading, but only what they are creating - and making a broad assumption that content creation somehow equates to experience. The more a person creates content relating to a concept, the more he/she is associated with that concept, hence the greater the indicator of experience (which may be roughly equated with expertise, etc.). So in principle, you can use e.g. Starmind technology to 'find an expert'.

Regards

Jonathan / UK


Matt Moore
 

Sam,

What problem are you solving?

Typically I do not find an expert for the sake of finding an expert. I find them because their expertise will help me solve a problem.

The kind of a expert that I need to find depends on the problem I am trying to solve.

Otherwise you find yourself in this situation.

“I’m not sure what to do with these answers”


Regards,

Matt Moore
+61 423 784 504

On Feb 24, 2021, at 1:11 AM, Sam Yip <sam@...> wrote:

Hi all - I am curious to understand the process by which you would identify a subject matter expert within your organization. I am working on a computer science research on identifying experts by reference to their information consumption pattern -- essentially looking at the topics of what one would read/write, and use that as a proxy to determine if he is an expert**. For example, if someone keeps reading and writing about "5G automation" and "carrier aggregation" (from his emails, blogs, documents, presentations etc.) then he is potentially an expert on these topics. Is this a simplistic way to approach a nuanced task? Do you have other processes to determine if someone is an expert? 

Looking forward to your thoughts

**the research methodology here is to apply algorithms to retrieve information from different channels (along with who writes/reads what), and detect pre-defined topics from the body of information.


Robert L. Bogue
 

Wow, those are big topics.

 

With regard to granularity the problem is a set of conflicting requirements.  If you make things too fine grained the complexity increases and the number of items per category drops – sometimes to irrelevance (one item.)  On the opposite side, if the taxonomy isn’t granular enough then you’ll end up with too many items in a single category and retrieval becomes difficult.  So the answer to your question is fundamentally about finding the balance between the opposing forces.

 

Experience / Expertise is illusive.  I wouldn’t try to capture except in the broadest scales (1-5) and then assume that it’s mostly wrong.   (ala Dunning Kruger Effect)  I’ve seen lots of attempts to capture experience/expertise and misses the point.  I don’t care how much you know about a topic if you have the answer that solves my problem.

 

For SME networks the balance is between power gradient and expertise.  Senior people will likely contact senior people first and the second person will do a referral to more specific experience.  Junior people will generally contact other juniors and mid-level people rarely reaching out to senior people because of the power gradient.  Of course culture has a strong impact on this.

 

Rob

 

-------------------

Robert L. Bogue

O: (317) 844-5310  M: (317) 506-4977 Blog: http://www.thorprojects.com/blog

Want to be confident about your change management efforts?  https://ConfidentChangeManagement.com

Are you burned out?  https://ExtinguishBurnout.com can help you get out of it (for free)

 

From: main@SIKM.groups.io <main@SIKM.groups.io> On Behalf Of Sam Yip via groups.io
Sent: Tuesday, February 23, 2021 10:02 PM
To: main@SIKM.groups.io
Subject: Re: [SIKM] Identify subject matter experts by information consumption?

 

Thank you all for the great comments and insights. A lot of valuable insights and there are a few points I’d love to learn more about:  

 

  • How do you determine the granularity of the taxonomy, which will be used as topics? In general is it always better to have fine-grained terms for taxonomy e.g. “SharePoint data migration” (fine-grained) vs “Information management” (very broad).  Also how do you strike the balance of letting taxonomy evolve on its own (which may be pulled in different directions by different offices across different regions) vs having a closely monitored and centralised taxonomy (at the risk of not keeping up with evolution of expertise/knowledge)?
  • it makes a lot of sense to differentiate experts - expertise - experience. How do you capture experience, or even just familiarity of a process, short of having a full-blown CoP or Q&A forum which requires a coordinated effort? To illustrate by way of example, say I am a project manager at a healthcare corporation at its Chicago office, and have previously looked into the process of applying for patents in Norway. I have spent 40 hours gathering information & requirements before the project was put on hold, but my effort will still be useful for for someone in say London office who is looking to apply for patents in Norway. How do you transfer that? 
  • In your SME networks, is there usually an inherent bias for senior staff, because they have built up sufficient expertise on topics? 


Tim Powell
 

The range of experience thoughtfully represented in this discussion is impressive!

 

Expertise Mapping and Location (EML) is a huge untapped resource in many organizations — true low-hanging fruit. As reflected in the comments here, it’s typically easier to envision than it is to execute.  There’s an ROI-focused EML case experience described in my new book, which is summarized in the attached article from Baseline magazine.  The payouts can be substantial.

 

The three main approaches being discussed here are:

(1) deducing expertise through scans of documents produced and/or sought;

(2) building databases/repositories of self-reported and/or assigned expertise/SME status; and

(3) building networks/CoPs that are largely self-defined and transactional (“Who knows about X?”).

 

In my experience, these as listed are in ascending order of effectiveness.  (2) and (3) can be combined, and this hybrid approach is probably the most powerful.

 

At best, this results in a sustainable internal “knowledge market” characterized by a vibrant ongoing exchange between knowledge users (i.e., seekers, “buyers”) and knowledge producers (i.e., providers, “sellers”).

 

The paradox I have observed is that whereas in most (non-knowledge) markets, the buyer pays — in knowledge markets, the seller too often “pays” through giving up his or her time/attention to provide the expertise.  The institutional challenge of providing sufficient rewards and/or recognition to render the effort self-sustaining is the hill that many of these effort stall on, in my experience.

 

Best,

 

Tim


TIM WOOD POWELL 
| President, The Knowledge Agency® Author, The Value of Knowledge

New York City, USA | DIRECT/MOBILE +1.212.243.1200 | ZOOM 212-243-1200

SITE www.KnowledgeAgency.com | BLOG www.KnowledgeValueChain.com

 

 

From: <main@SIKM.groups.io> on behalf of "Robert L. Bogue" <rbogue@...>
Reply-To: "main@SIKM.groups.io" <main@SIKM.groups.io>
Date: Wednesday, February 24, 2021 at 8:15 AM
To: "main@SIKM.groups.io" <main@SIKM.groups.io>
Subject: Re: [SIKM] Identify subject matter experts by information consumption?

 

Wow, those are big topics.

 

With regard to granularity the problem is a set of conflicting requirements.  If you make things too fine grained the complexity increases and the number of items per category drops – sometimes to irrelevance (one item.)  On the opposite side, if the taxonomy isn’t granular enough then you’ll end up with too many items in a single category and retrieval becomes difficult.  So the answer to your question is fundamentally about finding the balance between the opposing forces.

 

Experience / Expertise is illusive.  I wouldn’t try to capture except in the broadest scales (1-5) and then assume that it’s mostly wrong.   (ala Dunning Kruger Effect)  I’ve seen lots of attempts to capture experience/expertise and misses the point.  I don’t care how much you know about a topic if you have the answer that solves my problem.

 

For SME networks the balance is between power gradient and expertise.  Senior people will likely contact senior people first and the second person will do a referral to more specific experience.  Junior people will generally contact other juniors and mid-level people rarely reaching out to senior people because of the power gradient.  Of course culture has a strong impact on this.

 

Rob

 

-------------------

Robert L. Bogue

O: (317) 844-5310  M: (317) 506-4977 Blog: http://www.thorprojects.com/blog

Want to be confident about your change management efforts?  https://ConfidentChangeManagement.com

Are you burned out?  https://ExtinguishBurnout.com can help you get out of it (for free)

 

From: main@SIKM.groups.io <main@SIKM.groups.io> On Behalf Of Sam Yip via groups.io
Sent: Tuesday, February 23, 2021 10:02 PM
To: main@SIKM.groups.io
Subject: Re: [SIKM] Identify subject matter experts by information consumption?

 

Thank you all for the great comments and insights. A lot of valuable insights and there are a few points I’d love to learn more about:  

 

  • How do you determine the granularity of the taxonomy, which will be used as topics? In general is it always better to have fine-grained terms for taxonomy e.g. “SharePoint data migration” (fine-grained) vs “Information management” (very broad).  Also how do you strike the balance of letting taxonomy evolve on its own (which may be pulled in different directions by different offices across different regions) vs having a closely monitored and centralised taxonomy (at the risk of not keeping up with evolution of expertise/knowledge)?
  • it makes a lot of sense to differentiate experts - expertise - experience. How do you capture experience, or even just familiarity of a process, short of having a full-blown CoP or Q&A forum which requires a coordinated effort? To illustrate by way of example, say I am a project manager at a healthcare corporation at its Chicago office, and have previously looked into the process of applying for patents in Norway. I have spent 40 hours gathering information & requirements before the project was put on hold, but my effort will still be useful for for someone in say London office who is looking to apply for patents in Norway. How do you transfer that? 
  • In your SME networks, is there usually an inherent bias for senior staff, because they have built up sufficient expertise on topics? 


Valdis Krebs
 

Hi Sam,

Rather than looking at what an individual is doing, it is often more accurate to find out how others respond to the person. A true expert is someone who other's actually turn to for advice/expertise -- real behavior.  Not only do you want a real expert, but an expert who is willing to share his/her knowledge.  The best way to see all of this is via Organizational (Social) Network Analysis (ONA/SNA).  We can see who actually goes to whom for what expertise.  

Attached is a network map from an old client showing the organizations's expertise around the Java programming language.  Each node represents an employee (real names are hidden), a directed link ( ----> ) shows who goes to whom for expertise/advice on the Java programming language.  We can also measure such a network (see far right column of screenshot) so you get a map and a metric.

Valdis

Valdis Krebs

Orgnet, LLC

http://orgnet.com/about.html

 


 

Agree…ONA/SNA concepts and practice still valuable over the years Valdis..

 

Thanks

 

Bill

 

 

 

  

 

Learn more about the solutions and value we provide at www.workingknowledge-csp.com

 

 

 

 

From: main@SIKM.groups.io <main@SIKM.groups.io> On Behalf Of Valdis Krebs via groups.io
Sent: Wednesday, February 24, 2021 09:44
To: main@SIKM.groups.io
Subject: Re: [SIKM] Identify subject matter experts by information consumption? #expertise-location

 

Hi Sam,

Rather than looking at what an individual is doing, it is often more accurate to find out how others respond to the person. A true expert is someone who other's actually turn to for advice/expertise -- real behavior.  Not only do you want a real expert, but an expert who is willing to share his/her knowledge.  The best way to see all of this is via Organizational (Social) Network Analysis (ONA/SNA).  We can see who actually goes to whom for what expertise.  

Attached is a network map from an old client showing the organizations's expertise around the Java programming language.  Each node represents an employee (real names are hidden), a directed link ( ----> ) shows who goes to whom for expertise/advice on the Java programming language.  We can also measure such a network (see far right column of screenshot) so you get a map and a metric.

Valdis

Valdis Krebs

Orgnet, LLC

http://orgnet.com/about.html

 


Murray Jennex
 

My problem with a taxonomy for this is that taxonomies do not do a good job of relating like things.  They define a thing but not the things that are somewhat related.  Most of my issues with expertise is in the like things.  Searches show those that match the taxonomy but not those that are related and in engineering being too fine of a taxonomy makes it difficult to find the expertise you need.  This is why I tend to go more with the social network analysis approach to find those that people go to with questions.  Human minds and to some degree AI are much better at seeing the subtle relationships between topics than any taxonomy is at setting up a system to categorize expertise...murray jennex


-----Original Message-----
From: Robert L. Bogue <rbogue@...>
To: main@SIKM.groups.io <main@SIKM.groups.io>
Sent: Wed, Feb 24, 2021 5:14 am
Subject: Re: [SIKM] Identify subject matter experts by information consumption?

Wow, those are big topics.
 
With regard to granularity the problem is a set of conflicting requirements.  If you make things too fine grained the complexity increases and the number of items per category drops – sometimes to irrelevance (one item.)  On the opposite side, if the taxonomy isn’t granular enough then you’ll end up with too many items in a single category and retrieval becomes difficult.  So the answer to your question is fundamentally about finding the balance between the opposing forces.
 
Experience / Expertise is illusive.  I wouldn’t try to capture except in the broadest scales (1-5) and then assume that it’s mostly wrong.   (ala Dunning Kruger Effect)  I’ve seen lots of attempts to capture experience/expertise and misses the point.  I don’t care how much you know about a topic if you have the answer that solves my problem.
 
For SME networks the balance is between power gradient and expertise.  Senior people will likely contact senior people first and the second person will do a referral to more specific experience.  Junior people will generally contact other juniors and mid-level people rarely reaching out to senior people because of the power gradient.  Of course culture has a strong impact on this.
 
Rob
 
-------------------
Robert L. Bogue
O: (317) 844-5310  M: (317) 506-4977 Blog: http://www.thorprojects.com/blog
Want to be confident about your change management efforts?  https://ConfidentChangeManagement.com
Are you burned out?  https://ExtinguishBurnout.com can help you get out of it (for free)
 
From: main@SIKM.groups.io <main@SIKM.groups.io> On Behalf Of Sam Yip via groups.io
Sent: Tuesday, February 23, 2021 10:02 PM
To: main@SIKM.groups.io
Subject: Re: [SIKM] Identify subject matter experts by information consumption?
 
Thank you all for the great comments and insights. A lot of valuable insights and there are a few points I’d love to learn more about:  
 
  • How do you determine the granularity of the taxonomy, which will be used as topics? In general is it always better to have fine-grained terms for taxonomy e.g. “SharePoint data migration” (fine-grained) vs “Information management” (very broad).  Also how do you strike the balance of letting taxonomy evolve on its own (which may be pulled in different directions by different offices across different regions) vs having a closely monitored and centralised taxonomy (at the risk of not keeping up with evolution of expertise/knowledge)?
  • it makes a lot of sense to differentiate experts - expertise - experience. How do you capture experience, or even just familiarity of a process, short of having a full-blown CoP or Q&A forum which requires a coordinated effort? To illustrate by way of example, say I am a project manager at a healthcare corporation at its Chicago office, and have previously looked into the process of applying for patents in Norway. I have spent 40 hours gathering information & requirements before the project was put on hold, but my effort will still be useful for for someone in say London office who is looking to apply for patents in Norway. How do you transfer that? 
  • In your SME networks, is there usually an inherent bias for senior staff, because they have built up sufficient expertise on topics? 


Murray Jennex
 

an example of the below and the event that actually got me into KM: I had a problem at my nuclear plant with one of my heat exchangers.  The taxonomy used to classify equipment did it by material type or by system the component was in.  I searched and found no like issues so I spent 4 months doing a special engineering study in order to come up with a solution.  When I was done and rightfully proud of my work I was sharing with the engineer a row away in our cubicled room and it turns out he had solved the same problem on a heat exchanger in a different system with a slightly different material composition a year earlier that would have saved me a few months of time.  The problem was the taxonomy didn't see the relationships between the two components but when I talked engineer to engineer it was obvious.  Hence I think any taxonomy based system also needs a link system to expertise based on social network analysis or even self identified expertise.  Engineers don't have a problem with self identified expertise, if the self that identifies themselves as an expert doesn't match up we quickly move to the next expertise and the word goes out about the first person.....murray jennex


-----Original Message-----
From: Murray Jennex via groups.io <murphjen@...>
To: rbogue@... <rbogue@...>; main@SIKM.groups.io <main@SIKM.groups.io>
Sent: Wed, Feb 24, 2021 2:52 pm
Subject: Re: [SIKM] Identify subject matter experts by information consumption?

My problem with a taxonomy for this is that taxonomies do not do a good job of relating like things.  They define a thing but not the things that are somewhat related.  Most of my issues with expertise is in the like things.  Searches show those that match the taxonomy but not those that are related and in engineering being too fine of a taxonomy makes it difficult to find the expertise you need.  This is why I tend to go more with the social network analysis approach to find those that people go to with questions.  Human minds and to some degree AI are much better at seeing the subtle relationships between topics than any taxonomy is at setting up a system to categorize expertise...murray jennex


-----Original Message-----
From: Robert L. Bogue <rbogue@...>
To: main@SIKM.groups.io <main@SIKM.groups.io>
Sent: Wed, Feb 24, 2021 5:14 am
Subject: Re: [SIKM] Identify subject matter experts by information consumption?

Wow, those are big topics.
 
With regard to granularity the problem is a set of conflicting requirements.  If you make things too fine grained the complexity increases and the number of items per category drops – sometimes to irrelevance (one item.)  On the opposite side, if the taxonomy isn’t granular enough then you’ll end up with too many items in a single category and retrieval becomes difficult.  So the answer to your question is fundamentally about finding the balance between the opposing forces.
 
Experience / Expertise is illusive.  I wouldn’t try to capture except in the broadest scales (1-5) and then assume that it’s mostly wrong.   (ala Dunning Kruger Effect)  I’ve seen lots of attempts to capture experience/expertise and misses the point.  I don’t care how much you know about a topic if you have the answer that solves my problem.
 
For SME networks the balance is between power gradient and expertise.  Senior people will likely contact senior people first and the second person will do a referral to more specific experience.  Junior people will generally contact other juniors and mid-level people rarely reaching out to senior people because of the power gradient.  Of course culture has a strong impact on this.
 
Rob
 
-------------------
Robert L. Bogue
O: (317) 844-5310  M: (317) 506-4977 Blog: http://www.thorprojects.com/blog
Want to be confident about your change management efforts?  https://ConfidentChangeManagement.com
Are you burned out?  https://ExtinguishBurnout.com can help you get out of it (for free)
 
From: main@SIKM.groups.io <main@SIKM.groups.io> On Behalf Of Sam Yip via groups.io
Sent: Tuesday, February 23, 2021 10:02 PM
To: main@SIKM.groups.io
Subject: Re: [SIKM] Identify subject matter experts by information consumption?
 
Thank you all for the great comments and insights. A lot of valuable insights and there are a few points I’d love to learn more about:  
 
  • How do you determine the granularity of the taxonomy, which will be used as topics? In general is it always better to have fine-grained terms for taxonomy e.g. “SharePoint data migration” (fine-grained) vs “Information management” (very broad).  Also how do you strike the balance of letting taxonomy evolve on its own (which may be pulled in different directions by different offices across different regions) vs having a closely monitored and centralised taxonomy (at the risk of not keeping up with evolution of expertise/knowledge)?
  • it makes a lot of sense to differentiate experts - expertise - experience. How do you capture experience, or even just familiarity of a process, short of having a full-blown CoP or Q&A forum which requires a coordinated effort? To illustrate by way of example, say I am a project manager at a healthcare corporation at its Chicago office, and have previously looked into the process of applying for patents in Norway. I have spent 40 hours gathering information & requirements before the project was put on hold, but my effort will still be useful for for someone in say London office who is looking to apply for patents in Norway. How do you transfer that? 
  • In your SME networks, is there usually an inherent bias for senior staff, because they have built up sufficient expertise on topics? 


Murray Jennex
 

I still like using social network analysis over self reporting.  Its better to see who people actually talk to than to map those who say they are experts


-----Original Message-----
From: Tim Powell <tim.powell@...>
To: main@SIKM.groups.io <main@SIKM.groups.io>
Sent: Wed, Feb 24, 2021 5:36 am
Subject: Re: [SIKM] Identify subject matter experts by information consumption?

The range of experience thoughtfully represented in this discussion is impressive!
 
Expertise Mapping and Location (EML) is a huge untapped resource in many organizations — true low-hanging fruit. As reflected in the comments here, it’s typically easier to envision than it is to execute.  There’s an ROI-focused EML case experience described in my new book, which is summarized in the attached article from Baseline magazine.  The payouts can be substantial.
 
The three main approaches being discussed here are:
(1) deducing expertise through scans of documents produced and/or sought;
(2) building databases/repositories of self-reported and/or assigned expertise/SME status; and
(3) building networks/CoPs that are largely self-defined and transactional (“Who knows about X?”).
 
In my experience, these as listed are in ascending order of effectiveness.  (2) and (3) can be combined, and this hybrid approach is probably the most powerful.
 
At best, this results in a sustainable internal “knowledge market” characterized by a vibrant ongoing exchange between knowledge users (i.e., seekers, “buyers”) and knowledge producers (i.e., providers, “sellers”).
 
The paradox I have observed is that whereas in most (non-knowledge) markets, the buyer pays — in knowledge markets, the seller too often “pays” through giving up his or her time/attention to provide the expertise.  The institutional challenge of providing sufficient rewards and/or recognition to render the effort self-sustaining is the hill that many of these effort stall on, in my experience.
 
Best,
 
Tim

TIM WOOD POWELL 
| President, The Knowledge Agency® Author, The Value of Knowledge
New York City, USA | DIRECT/MOBILE +1.212.243.1200 | ZOOM 212-243-1200
 
 
From: <main@SIKM.groups.io> on behalf of "Robert L. Bogue" <rbogue@...>
Reply-To: "main@SIKM.groups.io" <main@SIKM.groups.io>
Date: Wednesday, February 24, 2021 at 8:15 AM
To: "main@SIKM.groups.io" <main@SIKM.groups.io>
Subject: Re: [SIKM] Identify subject matter experts by information consumption?
 
Wow, those are big topics.
 
With regard to granularity the problem is a set of conflicting requirements.  If you make things too fine grained the complexity increases and the number of items per category drops – sometimes to irrelevance (one item.)  On the opposite side, if the taxonomy isn’t granular enough then you’ll end up with too many items in a single category and retrieval becomes difficult.  So the answer to your question is fundamentally about finding the balance between the opposing forces.
 
Experience / Expertise is illusive.  I wouldn’t try to capture except in the broadest scales (1-5) and then assume that it’s mostly wrong.   (ala Dunning Kruger Effect)  I’ve seen lots of attempts to capture experience/expertise and misses the point.  I don’t care how much you know about a topic if you have the answer that solves my problem.
 
For SME networks the balance is between power gradient and expertise.  Senior people will likely contact senior people first and the second person will do a referral to more specific experience.  Junior people will generally contact other juniors and mid-level people rarely reaching out to senior people because of the power gradient.  Of course culture has a strong impact on this.
 
Rob
 
-------------------
Robert L. Bogue
O: (317) 844-5310  M: (317) 506-4977 Blog: http://www.thorprojects.com/blog
Want to be confident about your change management efforts?  https://ConfidentChangeManagement.com
Are you burned out?  https://ExtinguishBurnout.com can help you get out of it (for free)
 
From: main@SIKM.groups.io <main@SIKM.groups.io> On Behalf Of Sam Yip via groups.io
Sent: Tuesday, February 23, 2021 10:02 PM
To: main@SIKM.groups.io
Subject: Re: [SIKM] Identify subject matter experts by information consumption?
 
Thank you all for the great comments and insights. A lot of valuable insights and there are a few points I’d love to learn more about:  
 
  • How do you determine the granularity of the taxonomy, which will be used as topics? In general is it always better to have fine-grained terms for taxonomy e.g. “SharePoint data migration” (fine-grained) vs “Information management” (very broad).  Also how do you strike the balance of letting taxonomy evolve on its own (which may be pulled in different directions by different offices across different regions) vs having a closely monitored and centralised taxonomy (at the risk of not keeping up with evolution of expertise/knowledge)?
  • it makes a lot of sense to differentiate experts - expertise - experience. How do you capture experience, or even just familiarity of a process, short of having a full-blown CoP or Q&A forum which requires a coordinated effort? To illustrate by way of example, say I am a project manager at a healthcare corporation at its Chicago office, and have previously looked into the process of applying for patents in Norway. I have spent 40 hours gathering information & requirements before the project was put on hold, but my effort will still be useful for for someone in say London office who is looking to apply for patents in Norway. How do you transfer that? 
  • In your SME networks, is there usually an inherent bias for senior staff, because they have built up sufficient expertise on topics? 


Murray Jennex
 

totally agree Matt....murray


-----Original Message-----
From: Matt Moore <matt@...>
To: main@sikm.groups.io
Sent: Wed, Feb 24, 2021 2:18 am
Subject: Re: [SIKM] Identify subject matter experts by information consumption?

Sam,

What problem are you solving?

Typically I do not find an expert for the sake of finding an expert. I find them because their expertise will help me solve a problem.

The kind of a expert that I need to find depends on the problem I am trying to solve.

Otherwise you find yourself in this situation.

“I’m not sure what to do with these answers”


Regards,

Matt Moore
+61 423 784 504

On Feb 24, 2021, at 1:11 AM, Sam Yip <sam@...> wrote:

Hi all - I am curious to understand the process by which you would identify a subject matter expert within your organization. I am working on a computer science research on identifying experts by reference to their information consumption pattern -- essentially looking at the topics of what one would read/write, and use that as a proxy to determine if he is an expert**. For example, if someone keeps reading and writing about "5G automation" and "carrier aggregation" (from his emails, blogs, documents, presentations etc.) then he is potentially an expert on these topics. Is this a simplistic way to approach a nuanced task? Do you have other processes to determine if someone is an expert? 

Looking forward to your thoughts

**the research methodology here is to apply algorithms to retrieve information from different channels (along with who writes/reads what), and detect pre-defined topics from the body of information.