Creative Commons License
This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License. Knowledge is power, but kindness is greater

Konstantin M Golubev, Tatiana A Goloubeva
General Knowledge Machine Research Group

Kiev, Ukraine, E-mail: gkm-ekp at users.sf.net

Introducing a new kind of publishing:

The Electronic Knowledge Publishing

General Knowledge Machine Research Group White Paper

By Konstantin M Golubev, Tatiana A Goloubeva

Revised 9-Dec-1999, 1-Sep-2008

"The more original a discovery, the more obvious it seems afterwards"
(Arthur Koestler)

Publishers attacking

It is widely known what is it publishing. Publishers' products are around us all the day.

And sometimes it seems that they overproduce it. So why to speak about a new kind of publishing ? Isn?t it enough ?

Sorry for taking your time, but we would like to attract your attention to the questions of suitability of publishers production for solving tasks of real life and, in connection with it, to needs of a new kind of publishing introducing.

Knowledge and publishing of knowledge texts

There are, by our opinion, two main streams of publishing:

Emotional publishing has a goal to put people into some emotional state. In principle, there is no difference in this case between texts, pictures, sounds, movies, computer animations, and scents. The result should be emotional state in any case.

Knowledge, as we believe, is a basis of an ability of man to understand what is going and to act effectively in accordance with a situation.

Individual knowledge is an internal matter for a person, so no one needs, as a rule, to present this knowledge in external form for himself, such as text, picture etc. Sometime people use external objects to activate their knowledge, but as we think, if they have no knowledge inside already, nothing can be activated.

Knowledge existed long before any language was created. In fact, knowledge is one of the common features of living beings. A language was created partially for that purpose to pass individual knowledge to other persons, making this knowledge public and, if needed, everlasting.

Thus there were appeared external presentations of knowledge such as speech, texts, pictures etc. Primarily a speech and derived from speech text, that is a pictorial presentation of speech, are main modes of external presentation of knowledge. Therefore we can call this type of publishing of external presentation of knowledge - knowledge texts publishing.

We understand that this division of publishing on emotional and knowledge texts publishing is not very strict. The best emotional publishing, as a rule, contains some important knowledge about people, nature, and that makes it the most interesting. And many knowledge teachers try to present knowledge in a very attractive way.

And what is the problem with it?

The problem is that nobody can use effectively knowledge without making it his individual tacit knowledge with internal presentation. There were appeared such branches of human activity as a teaching and self-teaching. As a rule, teaching takes a long time and many efforts. Therefore we find that an impact of knowledge texts publishing on real life is strictly limited - a hard-working long-time teaching and learning is needed to get and use knowledge.

A human life is not long enough to learn all the knowledge even in a very special area. Many people say about information explosion. So what we can do? Treasure becomes larger with every day but for a whole life man can use just a little part of it.

There is no problem, of course, if we have a man near us who already possesses needed knowledge. We call him an expert. We just can describe a situation to him and get an advice how to act and what could be a possible result. The better case is if we can get advice from all the best experts and choose the better advice.

But it’s not easy thing to manage this consulting. As a rule, experts live in different places and have a very limited time for a communication. We know that it’s rare luck to get an advice from a good expert. And it?s almost impossible to gather together several good experts, especially in urgent cases.

Deadlock?

There were many efforts to bypass these problems. There were attempts to define a structure of knowledge texts, to introduce key words, hypertexts, databases, fuzzy search engines, Knowledge Management Suites etc. It makes easier to find knowledge texts regarding existing problems.

And in many simple cases it is working satisfactory. For example you can use a detailed instruction to maintain your TV set. It is interesting that even in this case before action a man needs to teach himself to make knowledge internal. But if a situation was not described in an instruction or a manual is very thick, you will call experts.

Other ways?

Yes. There were attempts for over 40 years to imitate an expert's activity. We can call works of W.Maccalloh and W.Pitts (Neural Networks); F.Rosenblat (perceptron); M.Minsky and S.Papert (anti-perceptronism, frames); Newell, Simon and Shaw (Logician-Theoretician, General Problem Solver); B.Buchanan and E.Feigenbaum (expert systems, DENDRAL, decision trees); E.Shortliffe (MYCIN, EMYCIN, deduction machine); H.Pople and J.Mayers (CADUCEUS); J.McCarthy (Artificial Intelligence, LISP); M. Ross Cuillian (semantic nets); R.Schank and R.Abelson (conceptual dependence, scripts, Script-Applier Mechanism, Memory Organisation Packets); A.Samuel (Self-teaching program); P.Winston (Arches); R.Michalski (self-programming, AQ11); G.Simon and P.Langley (programs for discoveries "Bacon"); D.Lenat (Automatic mathematician, Evrisco, RLL, CYC project); D.Hillis (Connection Machines, Thinking machines corporation); G.Hinton and S.Falmen (Boltsman machine) and many, many others. The most popular in this branch of research known as Artificial Intelligence (AI) are computer programs called expert systems and devices called neural networks. An expert system is an example of a 'top-down' approach when particular instances of intelligent behaviour selected and an attempt to design machines that can replicate that behaviour was made. A neural network is an example of 'bottom-up' approach when there is an attempt to study the biological mechanisms that underlie human intelligence and to build machines, which work on similar principles. There is interesting approach called Case Based Reasoning (CBR). It is applied to search similar cases in many help desk systems. Though it is not based on intellect simulation.

Were those attempts successful?

Sorry, but it is hard to say so. Some people say about a crisis of Artificial Intelligence.

But is this crisis of human intellect?

Of course, no. May be it’s a crisis of human self-confidence. In the beginning there were many promises to built machines more intelligent than people. And those machines should use advanced principles of work, much better than obsolete human intellect (see [4]). Instead of help to human intellect there were attempts to replace it. But those, who read works of academician V. Vernadsky from Ukraine ([1]), E. Le Roy ([2]) and P. Teilhard de Chardin from France ([3]), know that the main result of evolution on Earth is a creation of Noosphere - a sphere of intellect. And, in this case, it is very interesting what can be called an intellect, but is based on other principles than developed by evolution ?

So we believe that the real task is: to help human intellect, to make it more powerful and more creative, to let knowledge work for people using the principles developed by evolution.

And what are those principles?

How do you do it, Mr Sherlock Holmes ?

Let us look at the activity of the fiction's most famous detective Mr Sherlock Holmes. It is known that he has as prototypes real men: Dr. Joseph Bell of the Edinburgh Infirmary and Sir Arthur Conan Doyle himself (see preface [7]). So the methods of Mr Sherlock Holmes are absolutely realistic and widely recognized.

Every quotation below was drawn from [7]:

"No data yet... It is a capital mistake to theorize before you have all the evidence. It biases the judgement".

(Mr Sherlock Holmes)

A STUDY IN SCARLET, p.27

"I had... come to an entirely erroneous conclusion which shows, my dear Watson, how dangerous it always is to reason from insufficient data."

(Mr Sherlock Holmes)

THE ADVENTURE OF THE SPECKLED BAND, p.272

"For example, observation shows me that you have been to the Wigmore Street Post-Office this morning, but deduction lets me know that when there you dispatched a telegram...Observation tells me that you have a little reddish mould adhering to your instep. Just opposite the Wigmore Street Office they have taken up the pavement and thrown up some earth, which lies in such a way that it is difficult to avoid treading in it in entering. The earth is of this peculiar reddish tint which is found, as far as I know, nowhere else in the neighbourhood. So much is observation. "

(Mr Sherlock Holmes)

THE SIGN OF FOUR, p.91

Following Mr Sherlock Holmes, we can formulate the demand:

1. An expert should have a maximum possible information about a case before a judgement.

"You seem to be a walking calendar of crime", said Stamford with a laugh. You might start a paper on those lines. Call it the Police News of the Past?".

(to Mr Sherlock Holmes)

A STUDY IN SCARLET, p.18

"Kindly look her up in my index, Doctor,"... For many years he had adopted a system of docketing all paragraphs concerning men and things, so that it was difficult to name a subject or a person on which he could not at once fournish information"

(Mr Sherlock Holmes)

A SCANDAL IN BOHEMIA, p.165

"Like all other arts, the Science of Deduction and Analysis is one which can only be acquired by long and patient study, nor is life long enough to allow any mortal to attain the highest possible perfection in it."

(Mr Sherlock Holmes)

A STUDY IN SCARLET, p.23

Following Mr Sherlock Holmes, we can formulate the demand:

2. An expert should possess a maximum possible knowledge in a sphere of his activity.

"Now the skilful workman is very careful indeed as to what he takes into his brain-attic. He will have nothing but the tools which may help him in doing his work, but of this he has a large assortment, and all in the most perfect order." (Mr Sherlock Holmes)

A STUDY IN SCARLET, p.21

Following Mr Sherlock Holmes, we can formulate the demand:

3. An expert should possess no excessive knowledge, should have nothing but the tools which may help in doing work.

"As a rule, when I have heard some slight indication of the course of events, I am able to guide myself by the thousands of other similar cases which occur to my memory."

(Mr Sherlock Holmes)

THE READ-HEADED LEAGUE, p.176

"They lay all the evidence before me, and I am generally able, by the help of my knowledge of the history of crime, to set them straight. There is a strong family resemblance about misdeeds, and if you have all the details of a thousand at your finger ends, it is odd if you can?t unravel the thousand and first."

(Mr Sherlock Holmes)

A STUDY IN SCARLET, p.24

Following Mr Sherlock Holmes, we can formulate the demands:

4. Getting an indication of the course of events, an expert should be able to guide himself by the thousands of other similar cases which occur to his memory.

5. Possessing information about the thousands of cases, an expert should have an ability to find a strong family resemblance about them, i.e. to find templates of typical cases.

"As a rule,..., the more bizarre a thing is the less mysterious it proves to be. It is your commonplace, featureless crimes are really puzzling, just as a commonplace face is the most difficult to identify."

(Mr Sherlock Holmes)

THE READ-HEADED LEAGUE, p.183

"I have already explained to you that what is out of the common is usually a guide rather than a hindrance."

(Mr Sherlock Holmes)

A STUDY IN SCARLET, p.83

Following Mr Sherlock Holmes, we can formulate the demand:

6. An expert should focus on the most unusual in descriptions of situations.

"From long habit the train of thoughts ran so swiftly through my mind that I arrived at the conclusion without being conscious of intermediate steps. There were such steps, however. The train of reasoning ran, ?Here is a gentleman of a medical type, but with the air of a military man. Clearly an army doctor, then. He has just come from the tropics, for his face is dark, and that is not the natural tint of his skin, for his wrists are fair. He has undergone hardship and sickness, as his haggard face says clearly. His left arm has been injured. He holds it in a stiff and unnatural manner. Where in the tropics could an English army doctor have seen so much hardship and got his arm wounded? Clearly in Afghanistan.? The whole train of thought did not occupy a second."

(Mr Sherlock Holmes)

A STUDY IN SCARLET, p.24

Following Mr Sherlock Holmes, we can formulate the demands:

7. An expert should have an ability to explain the grounds of his conclusion.

8. An expert should arrive at the conclusion for a few seconds after getting a description of case.

"...you now pretend to deduce this knowledge I could only say what was the balance of probability. I did not at all expect to be so accurate."

(Mr Sherlock Holmes)

THE SIGN OF FOUR, p.93

Following Mr Sherlock Holmes, we can formulate the demand:

9. An expert should estimate a level of confidence of his propositions.

"In solving a problem of that sort, the grand thing is to be able to reason backward. That is very useful accomplishment, and a very useful one, but people do not practise it much. In the everyday affairs of life it is more useful to reason forward, and so the other comes to be neglected. There are fifty who can reason synthetically for one who can reason analytically...Most people, if you describe a train of events to them, will tell you what the result would be. They can put those events together in their minds, and argue from them that something will come to pass. There are few people, however, who, if you told them a result, would be able to evolve from their own consciousness what the steps were which led up to that result. This power is what I mean when I talk of reasoning backward, or analytically."

(Mr Sherlock Holmes)

A STUDY IN SCARLET, p.83

Following Mr Sherlock Holmes, we can formulate the demand:

10. An expert should have an ability to take into account not only descriptions of situations in his memory but results as well, providing a possibility to reconstruct a description from a result, i.e. if you told him a result, he would be able to evolve what the steps were which led up to that result.

"THE SCIENCE OF DEDUCTION"

A STUDY IN SCARLET, p.19

"For example, observation shows me that you have been to the Wigmore Street Post-Office this morning, but deduction lets me know that when there you dispatched a telegram... The rest is deduction...Why, of course I knew that you had not written a letter, since I sat opposite to you all morning. I see also in your open desk that you have a sheet of stamps and a thick bundle of postcards. What could you go into the post-office for, then, but to send a wire ? Eliminate all other factors, and the one which remains must be the truth."

(Mr Sherlock Holmes)

THE SIGN OF FOUR, p.91

Following Mr Sherlock Holmes, we can formulate the demand:

11. An expert should have an ability to point out all impossible hypothesises.

11 demands of Mr Sherlock Holmes

"He possesses two out of the three qualities necessary for the ideal detective. He has the power of observation and that of deduction. He is only wanting in knowledge, and that may come in time."

(Mr Sherlock Holmes)

THE SIGN OF FOUR, p.91

Following Mr Sherlock Holmes, we can describe steps of expert?s activity:

1. Observation

2. Producing propositions, based on a knowledge

3. Elimination of impossible propositions

4. Selection and verification of the most appropriate propositions

Thus, if we want to help human intellect, to make it more powerful and more creative, we should make a knowledge machine which could assist during these steps. Let?s name demands to such a machine.

Step 1 - Observation.

1. A knowledge machine should have a maximum possible information about a case before a judgement.

Step 2 - Producing propositions, based on a knowledge.

2. A knowledge machine should possess maximum possible knowledge in a sphere of it's implementation.

3. A knowledge machine should possess no excessive knowledge, should have nothing but the tools which may help in doing work.

4. Getting indication of the course of events, a knowledge machine should be able to guide itself by other similar cases which occur to it?s memory.

5. A knowledge machine should have an ability to take into account not only descriptions of situations in it?s memory but results as well, providing a possibility to reconstruct a description from a result, i.e. if you told it a result, it would be able to evolve what the steps were which led up to that result.

6. Possessing information about the great number of cases, a knowledge machine should have an ability to find a strong family resemblance about them, i.e. to find templates of typical cases.

7. A knowledge machine should have an ability to explain the grounds of it?s conclusion.

8. A knowledge machine should arrive at the conclusion for a few seconds after getting a description of case.

9. A knowledge machine should focus on the most unusual in descriptions of situations.

Step 3 - Elimination of impossible propositions

10. A knowledge machine should have an ability to point out all impossible propositions.

Step 4 - Selection and verification of the most appropriate propositions

11. A knowledge machine should estimate a level of a confidence of it?s propositions.

AI expert systems and neural networks

Let’s look at the Artificial Intelligence expert systems. Expert system, as we understand, is based on the idea of decision tree, when, with every answer to a program's question, a direction of moving through a tree changes until a final leaf (decision) will be reached (see [8]).

It is not amazing that development and implementation of a successful expert system is very hard work, because experts can not think, as a rule, in terms of decision trees, and the mathematical theory of probability have a little in common with a feeling of a confidence of an expert.

Let’s look at the Artificial Intelligence neural networks. Neural network is based, as we know, on the idea of teaching of set of elements (neurones), controlling conductivity between them (see [9]). Teaching is going under control of expert, which defines whether attempt is successful. This is more merciful towards expert - nobody is trying to make him feel himself deficient asking: what is the probability of this conclusion when that parameter's value is present. But there are some difficulties, not outdone yet.

Case Based Reasoning

CBR systems ability to retrieve related cases based on text matching is based on the statistical properties of the co-occurrence of 3-1 symbols grams. It means that frequently these systems presents as similar morphologically related cases that are of no semantic relevance. Though it is good general purpose searching tool.

Desicion making

The basis is the type of argument which is called 'uncertainty' in the literature of decision-making. This means that an argument is not clearly known and must be estimated, or as said in another way; one has to guess about the outcome in the future. This must be done by determining the behaviour in the past in similar cases, and extrapolate it to the future.

As we see, it is very close to Mr Sherlock Holmes demands.

Innovation Toolbox (Infinite Innovations Ltd)

There are two distinct parts of Innovation Toolbox:

1. The Probortunity solving section where you solve probortunities and come up with good ideas and store them in an Ideabase (a database of ideas).
2. A list of Problems, Opportunities and Good Ideas where you share problems to solve and share the good ideas you produce.

Building Knowledge Machine

And if we could build a knowledge machine satisfying 11 demands, it should mean that we could introduce a new kind of publishing - publishing of knowledge itself. Thus published knowledge could be used for tasks of real life with very few additional efforts.

How to build such a knowledge machine?

Famous experts in Artificial Intelligence (AI) Alan Newell and Herbert Simon, developers of General Problem Solver, proposed to define memory elements as rules called 'Productions' of the following type 'If Situation Then Action'. We have a right to suppose, taking into account this definition and opinion of Mr Sherlock Holmes, that big and important part of knowledge consists of following 3-parts elements:

(Description of real problem - Name - Action and Result), that is called a concrete knowledge, or (Description of template of problem - Name - Action and Result), that is called an abstract knowledge (we think that this kind of knowledge grows out of a concrete knowledge for a long lifetime).

We know that we can get information about knowledge from speech or texts. But they can be so amazingly long, ... just like this paper. So we should have to prepare a text of knowledge for input into a knowledge machine in a special way. How?

Let's consider the following example. There’s a need to develop a knowledge machine for pictures authors recognition assistance. We pick up a fragment from "Renaissance painting from Brueghel to El Greco" by Lionello Venturi ([5]).

"Like the Florentines, a Parma artist Francesco Mazzola (1503-1540), known as Parmigianino (i.e. little Parmesan), tended to the use of abstract forms, but, less doctrinaire in his abstractionism than such man as Rosso and Pontormo, he achieved a fragile grace and delicacy, reminiscent of Raphael and Corregio. His universal popularity contributed largely to the spread of mannerism in Europe.

The Madonna of the Long Neck (Uffizi, Florence) illustrates to perfection of his aesthetic. Here elegance replaces beauty and the somewhat abstract treatment of the figure gives it an immaterial charm. His sfumato, his discreet allusions to reality, the elongation of proportions and the sinuous movement of his figures were enthusiastically followed up by many painters in the second half of Cinquecento."

Let’s try to formalise it in a form (Description of template of problem - Name - Action and Result), because it is an abstract knowledge.

Description of template of problem

Description of problem consists of sentences that we call description signs. Every description sign, in principle, may have grades, usually five (why five - see [6]), for example (1 - Very Low, 2 - Low, 3 - Satisfactory, 4 - Good, 5 - Very Good), or have no grades at all. It is very important that every sign in description should be clear to any person and has one meaning. It is highly recommended to use sign in every description when it should be apparent.

In our case description of situation is set of ideas, derived from painter?s style description from previous fragment:

"Tendency to the use of abstract forms

Fragile grace and delicacy

Elegance replace beauty

Sfumato

Elongation of proportions

Sinuous movement of figures"

In our example, as we see, no description sign has grades.

Name

"Possible author is Francesco Mazzola (Parmigianino) from Parma, Italy (1503-1540)."

Action and Result.

Description of action consists of sequence of sentences that we call action signs. Every action sign describe a sequence of elementary actions called steps. Of course, there could be just one step.

In our case there's no action signs.

Description of result consists of sequence of one or more sentences that we call result sign.

In our case result sign could be:

"Possible author is Francesco Mazzola (Parmigianino) from Parma, Italy (1503-1540)." Since result is identical to the name here, in principle, it may be omitted.

Step 1 - Observation

Let’s look how (Step 1 - Observation) could be realised in a knowledge machine.

We gather all description signs, from all elements of knowledge that we have, eliminate synonyms and duplicates, and numerate signs and their grades. Signs numeration sequence doesn?t matter because numbers are just for a convenient reference. For a easiness of perception we can group signs any way, regardless of their numbers. As a result we get a chapter of problem's description input form called (1.Descriptions signs). The second part of problem's description input form chapter (2.Actions signs) consists of action signs, arranged in a way similar to description signs. The number of any action sign may not coincide with number of some description sign. The third part of input form chapter (3.Results signs) consists of result signs, arranged in a way similar to description signs. The number of result sign may not coincide with a number of some description sign or action sign. In principle, chapter (1.Descriptions signs) or (2.Actions signs), but not both, may not be present. Chapter (3.Results signs) may not be present at all.

So we get a form for input of description of problems with 3 chapters - (1.Descriptions signs), (2.Actions signs), (3.Results signs). For our example it should look like this:

Problem's Description Input Form

1.Descriptions signs

1.Tendency to the use of abstract forms.

2.Fragile grace and delicacy.

3.Elegance replace beauty.

4.Sfumato.

5.Elongation of proportions.

6.Sinuous movement of figures.

...

2.Actions signs

None

...

3.Results signs

200.Possible author is Francesco Mazzola (Parmigianino) from Parma, Italy (1503-1540).

...

In this form we should point out only that signs and grades, which are suitable for a description of existing problem.

Generated signs

In some cases there is a need to generate some useful signs using those which can be observed directly. Let's look at following quotation.

"The train of reasoning ran, ?Here is a gentleman of a medical type, but with the air of a military man. Clearly an army doctor, then. He has just come from the tropics, for his face is dark, and that is not the natural tint of his skin, for his wrists are fair. He has undergone hardship and sickness, as his haggard face says clearly. His left arm has been injured. He holds it in a stiff and unnatural manner. Where in the tropics could an English army doctor have seen so much hardship and got his arm wounded? Clearly in Afghanistan.? "

(Mr Sherlock Holmes)

A STUDY IN SCARLET, p.24

The observed signs are:

The generated signs (propositions) are:

Obviously, proposition itself could be a sign for further propositions. But we would like to remind that completely routine generation of propositions have to be tuned very carefully, otherwise it is not sufficiently reliable.

Step 2 - Producing propositions, based on a knowledge

Let?s look how (Step 2 - Producing propositions, based on a knowledge) could be realised in a knowledge machine.

Initially, we should numerate Name parts of knowledge elements, that will be used as propositions, just for convenient reference. It will look like:

All possible propositions names

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

1. Francesco Mazzola (Parmigianino) from Parma, Italy (1503-1540)

2. ...

3. ...

Every proposition is accompanied with a list of numbers of signs and grades from problem's description input form.

Getting the most possibly full description of problem, we could build a list of elements of knowledge with the most similar descriptions. It could be presented in a menu-like list of propositions, sorted according to indexes, which present value depended on degree of similarity. In our case of pictures authors recognition assistance it will look like:

The highly valuable propositions

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

Index Number Proposition

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

90 % 1 ) Francesco Mazzola (Parmigianino) from Parma, Italy (1503-1540)

...

It is very interesting question what could be an index. We think that there are many possible solutions, but we developed our own Proposition Value Index, based on idea of member of USSR Academy of Science M. N. Livanov from Russia that the essence of memory associations is a spatial-temporal coherence of narrow-band periodical oscillations of central neurones sets activity (see [11]).

Step 3 - Elimination of impossible propositions

Let's look how (Step 3 - Elimination of impossible propositions) could be realised in a knowledge machine. We know that knowledge element may have as a Name part a proposition like:

2.Parmigianino may not be an author.

Propositions of this type usually are very valuable, if Proposition Value Index is used. So there's a possibility to verify manually are there any objections against your favourite propositions ?

Step 4 - Selection and verification of the most appropriate propositions

Let's look how (Step 4 - Selection and verification of the most appropriate propositions) could be realised in a knowledge machine.

If you choose some proposition from list to get additional information, you should get a list of signs on which proposition is based. In our case, if we select the proposition:

90 % 1 Francesco Mazzola (Parmigianino) from Parma, Italy (1503-1540),

it should be like:

Proposition was made according to the following signs

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

4.Sfumato.

6.Sinuous movement of figures.

...

Next step - an additional information for more detailed verification of proposition should be present. It should be like:

Francesco Mazzola (Parmigianino) from Parma, Italy (1503-1540)

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

Tendency to the use of abstract forms

Fragile grace and delicacy

Elegance replace beauty

Sfumato

Elongation of proportions

Sinuous movement of figures

...

And you have a possibility for additional verification. After that you should return to a list of propositions.

The practical implementation

It was very interesting: is it possible to build a knowledge machine satisfying 11 demands of Mr Sherlock Holmes, and will it be really useful? Konstantin M Golubev and Tatiana A Golubeva have founded in 1986 General Knowledge Machine Research Group, based in Kiev, Ukraine. We are permanent staff of this informal group (our educational background is mathematics, theory of probability, graduated from Moscow and Kiev State Universities), variable staff are several information systems specialists and experts in different areas. We are not only mathematicians, but information systems experts as well. In 1987 the first version of General Knowledge Machine (GKM) was developed for RSX-11M operation system for Digital PDP-11 mini-computers. In 1989 second version of GKM was developed for UNIX operation system. In 1990-94 versions of GKM were developed for MS-DOS for IBM PC. In 1997 version of interface was developed for MS Windows 3.1/95/NT. In 1998 Internet version was developed for a Web server IIS 4. At the present time there are versions for any hardware and software platforms supported by GNU compilers.

In 1986-1990 we were working in a medical computer centre, so we have established good relations with medicine experts from different institutions, for example, from Institute of Gerontology, Kiev, Ukraine. We have made joint attempt to implement GKM. Results of this attempt were used in M.D. dissertations and were published in several joint papers. It was appeared, that results of implementation of Electronic Knowledge System (EKS), based on GKM as diagnostics support machine are very interesting. Tuning GKM, with the help of medical expert in radiography, we have checked almost 200 medical cases, comparing proven diagnosis with advice of GKM. It was appeared, that in absolute majority of cases the first one of the most valuable propositions of GKM was the same as proven diagnosis. In 1990-94, working on Parkinson's disease problem with medical experts from Institute of Gerontology, we used ability of GKM to find templates of typical cases (see [10]). In these articles we used this method called natural cluster analysis. It allows finding regularities while processing unstructured set of facts. It is very like a way a man finds typical cases among others and forms groups regarding these typical cases. In 1995-97 there was developed 'Electronic Knowledge System - Banking Management Strategy' for President of commercial bank ENERGOBANK, based on experience of banking community of Russia and Ukraine. In 1998 there was developed Internet version of 'Electronic Knowledge System: Negotiations - Language of Gestures'.

 Fights, threats and prospects

Will it be fight between knowledge texts publishing and Electronic Knowledge Publishing ?

No, we don’t think it should be. Knowledge texts publishing is for a knowledge learning. If you want to be an expert, you should read knowledge texts. Electronic Knowledge Publishing is for knowledge using. Electronic Knowledge Systems, as a rule, should be built on the basis of knowledge texts. We think the possible way might be electronic publication of knowledge, containing in the best knowledge books published already, in a form of Electronic Knowledge Systems (EKS) software or as on-line Internet network resource, and publishing of new knowledge books accompanied with GKM-based EKS software.

Is there any threat because of knowledge machines introducing ?

Some experts can have fears that EKS will be their competitors. It is obvious that the best and the most skilful users of knowledge machines will be experts themselves. Building, testing EKS are impossible without experts participation in some way. Just like only prominent expert is able to write a good book, only prominent expert is able to participate in building of a good knowledge machine. But it doesn't mean that the mechanism of creation of GKM-based system is difficult. An expert or experts team can build such a machine in a short time.

Prospects ?

The knowledge accumulated by people becomes larger with every year, and possibility to learn it remains the same due to limitations of a human brain. Therefore people become more specialized and have, for example, in case of medicine, less chances to evaluate state of patient. It leads to wrong or incomplete diagnostics and to ineffective treatment. Official list of World Health Organization contains tens thousand possible diagnoses. It is no surprise than no man in the world could learn and use them.

We think that prospects to apply knowledge of many people to problem solving, such as diagnostics, legal cases, management consulting etc., are attractive to all.

Our goals

We are looking for a co-operation in Electronic Knowledge Publishing introducing. Any arising propositions will be evaluated with gratitude. Take a look at our Internet Web Sites:

http://gkm-ekp.sf.net

Address: Konstantin M Golubev,

GKM Research Group coordinator,

E-mail: gkm-ekp at users.sf.net

 

About authors.

Konstantin M. Golubev,

Graduated from Moscow State University in 1977, mathematician, theory of probability and mathematical statistics.

Positions (through all years): mathematician, software engineer, lead programmer, vice-president of software development venture, chief of commercial bank Information Systems department, Information Systems consultant, since 1997 Head IT Expert at Space Research Institute, Kiev, Ukraine since 2002 CEO of consulting firm, coordinator of General Knowledge Machine Research Group.

Tatiana A. Goloubeva,

Graduated from Kiev State University in 1977, mathematician, theory of probability and mathematical statistics.

Positions (through all years): mathematician, software engineer, international databases access point administrator, public relations manager, e-mail and system administrator, S.W.I.F.T. System Administrator, since 1997 Computer Publishing Expert at National Bank of Ukraine, Public Relations Manager of General Knowledge Machine Research Group.

References (some text in Russian).

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

1. Vladimir I. Vernadsky. Biosphere.

2. E. Le Roy. Les origines humaines et l'evolution de l'intelligence.

3. P. Teilhard de Chardin. La place de l'homme dans la nature, R.,1981.

4. Roger Schank, Larry Hunter. The quest to understand thinking. Byte. By McGraw-Hill, Inc., New York, 1985.

5. Renaissance painting from Breughel to El Greco. Text by Lionello Venturi, Translated by Stuart Gilbert. Editions d'Art Albert Skira S.A., Geneva, 1979.

6. V. K. Oshe. A role of operative memory in solving of tasks of visual interpolation of linear intervals. In book "Psychophysiological regularities of perception and memory". Nauka, Moscow, Russia, 1985.

7. Sir Arthur Conan Doyle. The Penguin Complete Sherlock Holmes. With a preface of Christopher Morley. Penguin Books, 1981.

8. J.L.Alty and M.J.Coombs. Expert systems. Concepts and examples. The National Computing Centre Limited, 1984.

9. Geoffrey E. Hinton. Learning in parallel networks. Byte. By McGraw-Hill, Inc., New York, 1985.

10. G.N.Kryzhanovsky, N.B.Mankovsky, I.N.Karaban, S.V.Magaeva, N.A.Trekova, I.A.Vetrile, L.A.Basharova, M.A.Atadzhanov, K.M.Golubev. Serotonin antibodies and their possible role in Parkinsonism. Neurology and Psychiatry Magazine named after S.S.Korsakov, v.94., 1994, No. 5, Moscow, Russia.

11. Livanov M.N. Space organization of brain's processes, 1972, Nauka, Moscow, Russia)