The International Legal Technology Association advises when you “build” knowledge management : think windmill, not treadmill !
Treadmill meaning that your knowledge management needs human efforts to succeed, whereas a windmill is driven by the “wind” of your normal processes.
Or put otherwise : do not hire someone to “do the knowledge management” (~librarian !) but make it part of everyone’s job and even better : automate the information gathering and structuring, with textmining applications.

(Apparently, this “treadmill vs. windmill” concept was first set
forth in an article by Dan Felean of PensEra Knowledge Technologies, which
appeared in ILTA’s November, 2002 white paper titled “Knowledge
Management.” The full article can be found at http://www.iltanet.org under
“Communications,” “White Papers and Surveys.”)

According to Warmoes a knowledge management consultant firm, knowledge management is about getting answers to questions. Indeed that covers a major part of it. Interesting is their 30:30:3 rule :

• find an expert with the answer in 30 minutes
• give me a document with the answer in 30 seconds
• find the figures with the answer in 3 seconds

I wonder if someone can garantee these responsetimes. Although the first, if you can make the second work and if the expert is someone of your own enterprise should also be indentified in 30 seconds. What is the difference between a web page/resume and a document to find ?

What was first ?

When I read : “A ground-breaking new theory postulates that young children are able to learn large groups of words rapidly by data-mining” on Slashdot I start to wonder : did we not invent the datamining algorithms by copying the way an animal brain works (e.g. artificial neural networks) ? So why be surprised that if you copy something the original looks like the copy ?

Interesting post on swarm intelligence on the Center for Anarchic Strategy
A part from telling what swarm intelligence is all about there was this peculiar sentence : crowds tend to be wise only if individual members act responsibly and make their own decisions.  A group won’t be smart if its members imitate one another, slavishly follow fads, or wait for someone to tell them what to do.

Of course this makes sense for someone thinking about anarchy : no boss needed. This is also an important lesson for enterprise leaders or anyone in hierarchical positions : encourage your people to take their own decisions, do not be a boss telling them what to do or to behave as the others do, but ask them to take their responsabilities to help achieve the enterprise goals.

About a year ago Luis Suarez cited one post as, I quote, “of the most interesting and enlightening weblog posts I have seen in a while around the world of Knowledge Management” , end quote.
The post is called 15 Ways to Use Software to Improve Your Knowledge Management

I agree with Luis Suarez that it is indeed very interesting, giving a fairly good review of the softwares available.
However, the four goals that are distinguised as different goals look all about the same to me. After all : what is the difference between software to “encourage people to take advantage of other people’s knowledge“, to “ensure everyone can find the documents and other resources useful to them“, to “help staff easily answer common questions” or to “ensure senior staff has the right information to make decisions” ?
It all ends up to : “ensure that anyone easily gets the information he or she needs”.

As I wrote previously : you need i) to encourage people to share information, ii) have the means to store information, iii) ensure everyone can find the information and eventually use some tools to create information.

So what types/groups of tools do you need ?

The first group I call : make people communicate !
The best way to accomplish this is : put them together. They will communicate and need no software or tools at all to do this. This is wonderfully demonstrated by the google report on prediction markets (pdf) . The other traditional way is : telephone.
Then you have the more modern computer-based solutions : e-mail, blogs, instant messenging, social software.

The second group is : information storage. Let it be expertise repositories, simple file systems, knowledge bases, wiki’s, faq’s : first of all they should allow you to store any information you may need, but that is the simple part. The other parts are : ii)they should not only allow but even encourage people to put information in there : it has to be simple. And last but not least :iii) people should be able to easily find in a few clicks any information they are looking for.
And the trick here, and this I see as a critical success factor, is that people do not have to know where, or in which tool they have to search !
I often heard : “once you know where you can find it, its easy”. OK, but I want to find information anyway, even if I do not know where it is !
So : ONE starting point, one good SEARCH, but above all : one INFORMATION MODEL.

Howmany enterprises started knowledge management by modeling the information they needed ? If I should ask : howmany software development teams started writing software whithout doing some data- or objectmodeling, I am sure you will find it a ridiculous question : of cours none ! Is a knowledge management system not exactly the same, but even more complicated ?
So make an information model !
You may want to know which (sort of) information (a good old entity-relationship model is a good start) is needed by whom , when , howmany times etc… Otherwise the only clue to decide which software to use to enhance your knowledge management is your … software vendor.

                                     

I stumbled upon an article, describing some KM-models. Frankly I do not think this means a lot for people who want to do something about the knowledge in their enterprise. (Here I was at first going to write “to implement knowledge management”, but this seems way to abstract, and I assume in every enterprise there is already some sort of knowledge management, but not necessarily with that name).

What I am missing in that article is about the knowledge itself. The categories are not concrete enough. Yes, you find the distinction between tacit and explicit knowledge, but can somebody exactly explain when the first ends and the second begins ?

It seems to me like there is nowhere a description as where to put which information. What information should you put in an intranet ? What information should be shared in a classroom ? What information should be kept in documents ? What information should you put in a wiki ? etc…

here the article

Apparently swarm algorithms in datamining are still confined to universities. Although useful, e.g. for cleaning up digital images by means of particle swarm optimisation (Malik Braik and Alaa Sheta of the Department of Information Technology, at Al-Balqa Applied University) or the Antminer+ algoritm B.Baesens at the University of Leuven, Belgium), the practical use is severely limited by the computational efforts needed for the calculations.
Unless someone knows a scalable product, wich can easyly work its way trough a few hunderdthousends of records with 1,000 variables in an acceptable time ?

So you do not give it away just like that.

A few days ago Steven Burda asked a question on linkedin whether you should or should not withold any knowledge. Can you get an advantage from the fact that you do not share every work-related info with you colleagues ?

This question was followed by a whole lot of answers, ranging from ‘off course’ to ‘no way’, but the majority were in favor of keeping the most valuable knowledge nuggets for themselves. Which means that like in all biological systems there is no such thing as real altruism. We only give to be given.

Somewhere I read : “only fools are honest enough to tell the whole truth”.

But on the other hand, also purely egoïstic some people argued you shoud give away information to get new information in return: the win-win situation, which is very nicely illustraded by following finnish riddle by
Marko Peltojoki: What is the only thing which increases when divided/shared?
– Knowledge
Much depends on who you are working with, on the culture of your organisation.
Anyway, the question received 97 answers clearly showing that people recognise themselves in this (wanting to) withold information.
As I said : real altruism does not exist. Richard – the selfish gene – Dawkins already knew.

Previous posts dealt with learning and sharing, the two knowledge-related parts of knowledge management. The rest of knowledge management deals with information.

In my post about sharing I wrote about information creation when people share their knowledge.

There is yet anoter means of information creation : the automated one. I agree, computers do not do anything without humans telling them to (programming and the like), but still, the automated processes are capable of creating new information ready for humans to learn it and thus turn it into knowledge.

Those automated processes are

• data extraction/reporting : without these processes, databases would be useless as information sources for mankind. Hereby I for convenience I do not consider browsing (a database or text document) as reporting, but rather the computing-intense summarization, calculation of simple statistics, drawing graphs etc.
• data mining : looks for patters in existing data collections. To get a good idea of what it is, check KDnuggets
• text mining, sometimes labeled as mining of unstructured data (vs.datamining is mining of structured data). In reality it is composed of two steps : i) turning the unstructured data into structured data and ii) mining the structured data

In my previous post I wrote about learning, the ultimate goal of knowledge management.

Of course, in knowledge management, the opposite of learning is “sharing”. With learning, information is transformed into knowledge in someone’s head. With sharing, knowledge is gotten out of someone’s head and transformed into information, which can be used in …learning.

So here is a list of possibilities of knowledge sharing (or should I say information sharing?) :

• teach or give a presentation
• be a mentor
• write books, articles etc.
• publish on the internet (manage a website, web 1.0)
• put content on the internet (blog, write in wiki’s …, web 2.0)
• publish on the intranet (manage a website, web 1.0)
• put content on the intranet (blog, write in wiki’s …, web 2.0)
• add info to corporate database (is perhabs your job ?)
• speak at meetings, conferences
• help colleagues, explain things to them

Any suggestions for additions are welcome !

In a previous post on knowledge management I argue that information becomes knowledge only when it gets “between” the ears, when people get to know the info.
So this post deals with how to turn the information into knowledge, in other, not as fancy words : how kan we learn ?
There are a lot of ways and some people prefer one way, someone else prefers another way and still a third one prefers a combination. It is important to provide opportunity for all the ways of learning !
Here is the list

• in class
• from a mentor
• reading books, articles etc. : selfstudy on hardcopy
• internet surfing (googling) : electronic selfstudy
• web 2.0 : electronic selfstudy II
• intranet surfing (searching) : electronic selfstudy III
• enterprise 2.0 : electronic selfstudy IV
• querying in corporate databases
• attending meetings, conferences
• ask colleagues
• on the job training, which can be a mixture of anything

I’m not sure this list is exhaustive, so I will appreciate any other suggestions.

And just to think over : imagine a monkey in a scientific library : where is the knowledge ?

And a few more or less interesting links :

knowledge workers
Don’t Think Document Management, Think Knowledge Mgmt
What the heck is Knowledge Management? explaination – VIDEO

In an article called the nonsense of knowledge management  T.D. Wilson argues that about everyone uses the word “knowledge” in stead of “information”. Vendors did a “change all” in their manuals and called their software “knowledge management software”, but otherwise changed nothing. Try it. Take whatever text on knowledge management, do the great “change all” (replace “knowledge” by “information”) and see for yourself. Nothing happens.
I especially love his definition of knowledge : information between the ears.
In other words : information someone knows.
So forget about knowledge management and let us go back to do some proper information management in order to enhance the quality and quantity of knowledge in our enterprise.

Sometime ago I wrote something about psychons, a sort of elementary particles of the mind.
Today I discovered a taxonomy of psychons I made myself some three years ago. I did not know any more. Was quite amusing to discover it. It is not annotated, but should not need any explanation.

Since I began to search about knowledge management some months ago it became clear that first there is al lot more about it than just a document database and second that no single source covers it all.

I want to make a sort of complete powerpoint presentation on the subject, so I first have to find some order in the mess of subjects that have to do with … yes , the subject.

A mess ? Yes indeed. What follows is that heap of subjects I found. Until now I did not bother to sort them out. That what’s coming in the upcoming parts.

If you are curious, search a bit in the heap.

All comments, suggestions and other positive feedback is extremely welcome.

The heapAn information retrieval system will tend not to be used whenever it is more painful and troublesome for a customer to have information rather than for him not to have it
service oriënted architecture and KM
share ideas, reward the accepted ideas, prediction market on the ideas
idem for knowledge bits, wiki pages
all the different ways of learning (classical, internet search, mentor, colleague’s…)
a “who to ask” database
Quality of knowledge versus availability. Availability first !!
Culture/ cash /
Do’n t impose a too rigid structure
Gap analysis to know where knowledge is missing or not accessible
will sharing bring me more money than keeping it ?
Do you sell it or do you give it away ?
put “sharing” (yearly) objectives/evaluations
“Figures => data=>info=>knowledge=>wisdom=>sharing
Each colleague studies one topic a year (a part from his normal work) and shares the info with the others (presentation)
Anyone can suggest ideas for the above
KQM needed: knowledge Quality manager to make sure that the kwowledge base does not become a garbage can (eg. titles as “info” and the like prevent info in twofold …)
“Once upon a time, saying the words ?knowledge management? was the fastest way to get thrown out of a meeting. And not only would you be ejected, you would never be invited back.”
Cynefin :

• simple domain

Few cause-effect relationships, simply connected, so that cause and effect connections are predictable and repeatable. Problems that are solved tend to stay solved, at least in the medium-term. Bureaucratic structures work as a way to organise action, because best practice can be built into procedures. The learning sequence is sense-categorise-respond.

• Complicated domain

Multiple causes and effects, where cause-effect relationships can’t be immediately recognised, but can be understood through research. Professional structures handle this domain best, using communities of practice to evolve and confirm best practice. The preferred learning sequence is sense-analyse-respond.

• Complex domain

Multiple causes and effects with complex feedback loops. Causes are only perceivable after the end, and work as explanations rather than predictions. Problems keep returning despite many attempts to resolve them, and solutions imported from elsewhere don’t work. A lot can be learned from mistakes. Collegial structures work best, and the learning sequence is probe-sense-respond.

• Chaotic domain

Multiple and turbulent cause-effect connections, so turbulent that there is no point in thinking about cause and effect. There is no point in trying to develop solutions: it’s best to act quickly to create stability. Any structure is of little help: just act-sense-respond.

“strongest correlation in betting was found among people who sat very close to one another, trumping even friendship or other close social ties…. This is tangible evidence, the authors argue, that information is shared most easily and effectively among office neighbors, even at an Internet company where instant messaging and e-mail are generally preferred to face-to-face discussion.”

So much for working virtually or formal knowledge management. Maybe it does all come down to proximity! Or perhaps workers waste time in idle chit chat with neighbors. Take away the neighbors and then they have to communicate with actual co-workers. Of course, if you can group workers in close physical proximity, that indeed has value. But given global teams, that often is just not possible.

Knowledge ~data model
Management ~process model
A complete analysis is necessary of all the processes that generate info and also need info
Data smog

So far the heap

These two items are to happen in this very century :

1. The north pole is dissapearing. When will we have a summer with no north pole any more? Some say as early as 2040.
2. A human-level artificial intelligence is on it’s way. Ben Goerzel on kurzweilai.net believes it will take mankind only 10 years to create an artificial intelligence that maches a human brain. But Ray Kurzweil himself thinks it will take twice that time.

Anyway : it is for sure that there are interesting times coming up but also that humans are radically changing the world.
As for the global warming and dissapearing of the north pole : if the ice of Greenland continues melting soon or later the gulf stream will stop (not everyone agrees on this) and cause a  drop in temperature in northwestern Europe. Yes, global warming will cause a temperature drop !
And the human artificial intelligence will help humans create a superhuman artificial intelligence that will continuously improve itself : it wil become a singularity.
What this will do to us ? no one knows ! As long as we can pull the plug and shut it down we could keep feeling pretty safe, but will we pull all plugs ??
At those times I will be an old man if I am lucky to live long enough, and whatever comes, so what. But what about my children and their children ? Pretty scary, no ?

In the posts behind this link collective intelligence is discussed. Interesting, and as I like structure, a short list of three points has attracted my attention :

• collective reflexion : thinking together
• collective communication : well, since in communication allways at least two parties
  are involved, we can simply drop the “collective”
• decision making : can be enhanced by information that comes from “collective” communication.

Here the blogger clearly missed his start. I would start to wonder : what is intelligence ?
Wikipedia : “… capacities to reason, to plan, to solve problems, to think abstractly, to comprehend ideas, to use language, and to learn…”
Simply put : the capacity to absorb information and to do something useful with it.
So what about “collective” ?.
It is clear that if one individual is busy being intelligent he uses the info at hand, so alone, he is limited to his own information. He can do better by using information of others too : texts, pictures or sounds in books, dvd’s, on the internet (asynchonous), or what other people say or show directly to him alive or at television or msn (synchronous). Doing that it is already collective intelligence.
So if you want to implement enterprise 2.0 to get a higher collective intelligence, it’s all about sharing information, which leads us to the other buzzword : knowledge management.

Leila Summa gives 10 buzzwords for 2008 on her blog. Three of them which I am interested in are :

1. 2.0-itis
2. swarm intelligence
3. emergent systems and self organisation

For obvious reasons she gives no comment on the first.

The second she sees it als ‘intelligence of the mass’. Somewhat cryptically :”Kind of “intelligence of the mass” effects happens all the time and are hidden, but very important part of our lives – even we don’t realize it”.
I think this needs some elaboration.
In this explanation, intelligence of the mass, at least what concerns our human species, is not something of web 2.0, but something that is there all the time.
But what is swarm intelligence exactly ??
According to wikipedia it is “artificial intelligence …”
I suppose there is a lot of swarm intelligence in an ant heap, but no artificial intelligence whatsoever. So far for the definition by wikipedia.
hereI found a definition which is much closer tho what I understand under the term swarm intelligence :
“Swarm Intelligence (SI) is the property of a system whereby the collective behaviours of (unsophisticated) agents interacting locally with their environment cause coherent functional global patterns to emerge.”
Which means : agents do things that concern themselves, not knowing that they are part of a bigger picture. There are two distinct levels of knowledge.
Not knowing ??
Perhaps to be or not to be swarm intelligent is not so binary at all. Can we be a little swarm intelligent ? Meaning that either there is a little bit a bigger picture, or there is a bigger picture and the individual agents do know it or the agents do only loosely belong to the swarm … ?
In that sense I already had a discussion on this weblog wether or not stock traders show swarm intelligence.

And in a post yet to come I will write about collective intelligence.

On Emergent systems and self organisation she only ask questions, e.g. “When does a system start to be “emergent” or “self organized”?

Is a bit like the above discussion : when you can discern two levels of knowledge/intelligence/patterns.

But what if those two levels are not as neatly separated as in an ant heap ? What if they gradually flow over from the lower to the upper level ?

It’s all a bit like fuzzy logic, and it’s saturday evening, so I am going to pour for myself a good glass of french red wine. Cheers.

Novamente plans to release a flock of artificially intelligent virtual parrots into the online virtual worlds. Since their artifical general intelligent systems need teachers to learn, they want to use al the humans in the virtual worlds (the wisdom of the crowds!) to interact with the virtual parrots in order to learn english. And of course, after that they want to go further.

Read the story at the novamente weblog.

A lot of years ago, a collegue of mine and I were brainstorming for fun about mindreading and that sort of stuff and came to the concept of psychons : elementary particles of thought which can move from one mind to another.

Years later, I had not seen this colleague for years, we came in contact again and I remembered the psychons. But the difference now was : there was the internet. So I searched for “psychons”…..

Great was my astonisment when I discovered that at about the same time of our joking about psychons, there was a former nobel prize winner who published a similar idea in a respected scientific journal and in a book (here the Book review by Joseph Uphoff). Although he did not speak about telepathy or so, for him the psychons were the parts of the brain were the ‘self’ was situated. A lot of people criticised this idea (perform a google search on “psychons brain eccles” and you will see) for being religious oriented : mind-body dualism.
I will not dig further in the matter, only, at the time we considered psychons as a joke and my opinion on this did never change, not even for a nobel prize winner 🙂

If you enjoyed this post, then you might also be interested in the following : A taxonomy of psychons Web 5.0 : the telepathic web did dragons exist ?

 

                                     

Andrew Gent states that web 2.0 lacks “process”.
Is that so ? I think web 2.0 has enough processes.
He compares a wiki with an e-mail tool. an instant messenger, a web site. The differences he points out, are truly there. so far he’s right. But the difference is not that web 2.0 does not possess processes. The processes are different. For example In stead of sending a message, in web 2.0 we broadcast a message to whomever wants to listen or read. In web 2.0 we still publish our opinion, only the process is much more simple : there is no censoring going on.
But are these differences not exactly what Web 2.0 gives his identity ? Without these differences nobody would have called it “Web 2.0 ” !
Exactly this lack of hierarchy, these possibilities of anarchy (creativity is perhaps a better word?) make it web 2.0 and gives it all this new possibilities.
In a previous post I wrote about the difficulties management of an enterprise can have to adopt an Enterprise 2.0 culture.
I think again the pendulum of new technologies is swinging out a bit too far to be directly usable in an enterprise environment. And there I agree with Andrew that enterprises will have to find a good in-between between the old intranet, e-mail etc. on the one hand and the anarchical web 2.0 on the other hand.

Interesting question and dito answers in Linkedin : What is the Return on Investment of enterprise 2.0 ?
A lot of people say that ROI is something of enterprise 1.0 en should therefore not been used. Silly of course, you still have to make money somehow !
Others clearly state that enterprise 2.0 is a box of tools that can help in your project or whatever you do and therefore you should be able to see an increase in ROI of your project.
And at last Sebastien Wiertz hits the bullseye by saying that E2.0 is not only the toolbox but also a change in behaviour (culture ?) of the company. “you need to create a real relationship between the user of a team before being able to get them in a virtual Environment“

I think not every enterprise is ready for this. Perhaps you can begin with a department where a lot of the web 2.0 generation is working, where a lot of them (as I do) already think : if we only had those tools on the workspace! When that becomes a success story, time may be right to spread the word and the technology to the rest of the company.

This book (the title of this post is the title of the book) by Ajith Abraham, Crina Grosan, Vitorino Ramos (eds.), Springer-Verlag Berlin Heidelberg 2006, is an interesting compilation of chapters on swarm intellingence in data mining.
Contains a lot of different aspects of the swarm approach to solve data mining problems, like classification, feature selection, learning fuzzy rules etc.

Emergent patterns are the relatively complex patterns that we observe on a higher level and which are the result of relatively simple behaviour of agents of some sort on a lower level. Examples are bee swarms, ant hills, termite hills, animal brains, stock prices, multicellular organisms, human organisations etc..

For patterns to emerge, there are two conditions :

• first : you need to have a lot of individual agents at the lower level which belong somehow to an identifiable group
• second : you need a a whole lot of these indentifiable groups (the higher level)  in order to permit some sort of natural selection of the fittest groups.

Example : a termite hill is built by thousends of thermites. If their behavioral patterns are suboptimal, something will go wrong with the termite hill (e.g. bad air conditioning) and they all will die. The other termite hills who do better will survive and pass on their genes to the next generations.

WHEN DOES THIS EMERGENCE ENDS ?

It is simple : when one of the two conditions are no longer met.

Let us look at ourselves : humanity. We are the result of a long series of emergent patterns, beginning by organic molecules organizing into cells, by cells organizing into first simple multicellular organisms and later on into more complex organisms. We became intelligent and organised ourselves into communities (tribes) with various social and hierachical structures. The most successful of these communities have spread and taken over the whole planet.

So where do we stand as to new emerging patterns ? There are still a lot of these communities so the first condition for emerging patterns is fulfilled. But what about the second one ? Howmany groups of huge numbers of communities are there ? It seems as there is only one group left at the highest level. The communities and individuals of this group are connected through the internet and other telecommunication media. So there is no room for competition at the highest level and we humans have reached the end of emergence.

Do we ?

Eventually we will swarm out to the stars and each occupied planet will represent a different community, isolated enough from the other ones.  This could open up a new higher level of patterns.

Here is a useful references website on “swarming”.
Great for anyone interested in how swarming works.

Drawback : there are no categories dealing with swarm intelligence and artificial intelligence or data mining.

In an interesting paper on baysian statistics (”a technical explanation of technical explanation“) Eliezer Yudkowsky radically destroys the “emergence” hype. I follow his reasoning when he says that everything is called “emerging” without any logical reason. What does the word “emergent” means in sentences like (the sentences are his‘):

• human intelligence is an emergent product of neurons firing
• the behavior of an ant colony is the emergent outcome of the interactions of many individual ant

It is simple : the word “emergent” in these sentences has no meaning ! You can eliminate the word and the sentences do not lose any significance.

Moreover : in some other sentences you can even replace the word with another nothing-saying word : “magical”

• life is an emergent phenomenon => life is a magical phenomenon
• human intelligence is an emergent product of neurons firing => human intelligence is a magical product of neurons firing

I did a little test and typed “emerging patterns” in google. As a result I got things as

• Efficient Mining of Emerging Patterns: Discovering Trends and Differences
• We provide an executable file for users to discover emerging patterns from binarized data files
• Classification by aggregating emerging patterns
• Mining emergent substrings
• instance-based classification by emerging patterns

In each one of these and many other uses of the word, you can just eliminate it without consequence for the meaning of the sentences.

And especially in data mining (I am a data miner myself) you should not use the e*-word. Yes, data mining is about finding patterns in data. But the patterns do not “emerge”. They just are there in the data and it’s up to you to find them, to do the hard work of sampling, calculating, quality checking etc…

So let’s finish by saying that swarm intelligence is the result of patterns in the behaviour of for example individual bees in a swarm. Did you miss the e*-word ?