THE WELCH COMPANY
440 Davis Court #1602
San Francisco, CA 94111-2496
415 781 5700
rodwelch@pacbell.net


S U M M A R Y


DIARY: April 5, 2000 00:31 AM Wednesday; Rod Welch

Colloquium considers SDS for defining DKR environment.

1...Summary/Objective
2...Meaning Drift Identifies Opportunity for DKR Solution
3...Feedback Critical Metric of Communication for Effective Intelligence
4...Journal Can Be Considered for Application to DKR
5...Knowledge Representation Requires Use-case Analysis
6...Define Opinion, Analysis, Daily Working Information, Knowledge
7...Experiential Record, Like SDS, Useful for Knowledge Management
8...Journal Incorporated into DKR because Email is a Diary of Interactions
9...Email is a Form of Diary that Can be Incorporated into DKR
10...DKR Can Incorporate Personal Journal Since Email is a Form of Diary
......Links in SDS Harder to Follow than Inline Information
......SDS More Difficult to Follow Using Links than with Inline
......Inline Material Reduces Problem of Links Improves Productivity
..........Knowledge Management Dilemma Resolved by Better Technology
..........Intelligence Mind Boggling to the Conscious Span of Attention
......Competing Versions of Record are Messy When Many Use SDS
......Life Full of Competing Versions SDS Enables Early Discovery
......Discovery Conflict Goal of Intelligence to Avoid Mistakes
11...Knowledge Mandala Provides Five (5) Major Components of Knowledge
12...Nancy Glock-Grueneich Formulates Parts of Knowledge Representation
13...Welch Formulates Process for Applying "Meaning" Component Knowledge
....Ontology Needed for Analytical Forms, Systems Dynamics
....Meaning From Experience Supported by Welch's SDS Program
....Case Studies Derivitive of SDS May Support KR "Cases"
....Use Cases for Analytic, System, Meaning Knowledge Representation
....Order of Discussion Knowledge Representation
14...XML Helpful, But No Easy Solution for KM Design
15...Knowledge Representation Difficult Problem, No Industry Answer Yet
.....Representations Change Over Time Hard to Program Flexibility
.....Need Strong Solution to Easy Problem That Improves Productivity
16...Knowledge Representation Classes Enumerated

ACTION ITEMS.................. Click here to comment!

1...Need example of an ontology that uses "signs" and "symbols."
2...Need definition of "knowledge" per questions on SDS...
3...Paul proposes getting an overview of AI and knowledge management

CONTACTS 

SUBJECTS
Microsoft/Intel Dominate Market Share
Alignment Avoid Meaning Drift
Meaning Drift Avoid Alignment
Market Conditions
Communication Biggest Risk in Enterprise, Dilemma
Align Communications to Maintain Shared Meaning from Meetings, Calls
Align Communications, Testing Metrics
Align Meeting Understandings Feedback Metrics, Notes
Meaning Drift, Com Biggest Risk Enterprise

1411 -
1411 -    ..
1412 - Summary/Objective
1413 -
141301 - Follow up ref SDS 72 0748, ref SDS 71 0748.
141302 -
141303 - Answered question about original purpose of the Lifestreams subject,
141304 - noting "meaning drift" that makes communication the biggest risk in
141305 - enterprise, which presents an opportunity for the DKR. ref SDS 0 1680
141306 - Jack responded that his letters should not be published on the web
141307 - showing the names of people who get copies.  I fixed this. ref SDS 0
141308 - 6370  Jack submits a plan to develop Knowledge Representation.
141309 - ref SDS 0 5833  He suggests the Welch system of using experience
141310 - supports Knowledge Representation for meaning constructs. ref SDS 0
141311 - 0924  Jack requests opinion from the project team on using a journal
141312 - for the DKR, and asks, how "opinion" should be represented.
141313 - ref SDS 0 2928  Called Jack and discussed defining "knowledge,"
141314 - "opinion" and "daily working information" to guide design of DKR.
141315 - ref SDS 0 0448 Bill Beardon indicated a journal might be useful.  Eric
141316 - Armstrong commented a journal is not harmful, but the value is not
141317 - settled, and notes that an email forum is a kind of "diary."
141318 - ref SDS 0 1972  He proposes using "inline" rather than linking to
141319 - improve Rod's diary. ref SDS 0 4823 Paul Fernhout offers opinion on
141320 - XML for the DKR. ref SDS 0 5977  He says knowledge representations are
141321 - difficult to figure out because things change over time, ref SDS 0
141322 - 3600, and gives examples to illustrate this point, some of which seem
141323 - like traditional DBMS work, so focus on defining objectives of DKR may
141324 - be helpful. ref SDS 0 3705
141325 -
141326 -     [On 000406 sent follow up to Jack. ref SDS 75 0001
141327 -
141328 -
141329 -
141330 -
141332 -  ..
1414 -
1415 -
1416 - Progress
141701 -  ..
141702 - Meaning Drift Identifies Opportunity for DKR Solution
141703 -
141704 - Submitted ref DIT 1 0001 to Jack responding to his letter on 000403,
141705 - ref DRP 1 0001 asking about the objective of letters being submitted
141706 - by the Colloquium under the subject of Lifestreams. ref SDS 72 0748
141707 -
141708 - Jack's question about original subject, points to opportunity for SRI
141709 - project to solve meaning drift that makes communication the biggest
141710 - risk in enterprise. ref DIT 1 0870
141711 -
141712 -    Cited Jack's original letter on 000331. ref SDS 71 0748
141714 -     ..
141715 -    Cite my response on 000331. ref DIP 3 0001
141716 -    ..
141717 -    Drucker's article on 991025. ref SDS 32 0836
141719 -     ..
141720 -    Meaning drift from POIMS. ref OF 3 8316
141722 -     ..
141723 -    Explain as similar to being confused, on 991101. ref SDS 33 0857
141725 -     ..
141726 -    High cost of medical mistakes on 990924. ref DIP 2 0001
141728 -  ..
141729 - Mention efforts to enhance confidentiality reported on 000329,
141730 - ref SDS 68 0001, in relation to Jack's request on 000403. ref SDS 72
141731 - 0001
141732 -
141733 -
141735 -  ..
1418 -
1419 - 0922 received letter
1420 -
142001 - Received ref DRT 1 0001 from Jack responding to my letter sent this
142002 - morning, per above. ref SDS 0 1680
142004 -  ..
142005 - Jack does not address in his letter the original subject of Microsoft
142006 - vulnerabilities nor the subject I raised about meaning drift, as an
142007 - opportunity for new technology to be successful.
142009 -  ..
142010 - Jack cites concern of DKR that shows copies of letters issued by
142011 - others, and in particular by him. ref DRT 1 3780
142013 -  ..
142014 - I removed from the published record references to copies that are in
142015 - Jack's letters.
142017 -  ..
142018 - Confidentiality and privacy were considered in discussion with Jack on
142019 - 000331. ref SDS 71 3245
142020 -
142021 -
142023 -  ..
1421 -
1422 -
1423 - 1027 called Jack
1424 -
142401 - Jack seemed to indicate the original subject of his letter on
142402 - Lifestreams was not Microsoft Vulnerabilities, but rather merely to
142403 - point out an important publication by Geletner, whom he regards as a
142404 - leading voice on the future of technology, based on past
142405 - accomplishments.
142407 -  ..
142408 - Jack did not discuss analysis of Geletner's article in the record on
142409 - 000331. ref SDS 71 6887
142411 -  ..
142412 - We briefly, reviewed his letter on 000331 commending the Lifestreams
142413 - article. ref SDS 71 0748
142415 -  ..
142416 - We looked at Geletner's article titled...
142417 -
142418 -             Now That the PC Is Dead...
142419 -
142420 -             Microsoft needs no vanquishing by government lawyers;
142421 -             markets will take care of that
142422 -
142423 - ...shown at ref OF 1 0001
142425 -  ..
142426 - Jack seemed to indicate in the discussion that the opportunity to
142427 - solve meaning drift, cited in the letter, ref DIT 1 2457, is an
142428 - important objective of knowledge management, developed on 000403.
142429 - ref SDS 72 4814
142431 -  ..
142432 - Jack did not discuss why he did not mention this in his letter today.
142434 -  ..
142435 - Jack indicated frustration about not having greater command of the SDS
142436 - record on the Internet.  He gave an example of wanting the ability to
142437 - look up everything about "Jack."
142439 -  ..
142440 - We considered this capability is in the SDS program.  It requires a
142441 - variety of methods, including organic subject structure, what Jack has
142442 - called an "ontology."  The Colloquium is gaining experience to guide
142443 - development of these capabilities, which Jack cited in his letter on
142444 - 000331. ref SDS 71 2173
142445 -
142446 -
142447 -
142448 -
142449 -
142450 -
142451 -
1425 -

SUBJECTS
Diary Experience Integrate Documents
Chronology, Cause and Effect, Sequence Complexity
Traceability to Original Sources Align with Links to History Context
SDS Evaluation Slick Links Self-evident Benefits Innovation Dynamic
SDS Evaluation Organizational Mmeory Report Intelligence Knowledge S

5707 -
570801 -  ..
570802 - Feedback Critical Metric of Communication for Effective Intelligence
570803 - Journal Can Be Considered for Application to DKR
570804 -
570805 - Jack's first letter this morning discusses the Welch system of using
570806 - an experiential record to construct "meaning" that makes sense of
570807 - complexity in relation to objectives, requirements, and commitments,
570808 - which is part of Knowledge Representation, reviewed below. ref SDS 0
570809 - 0924
570811 -             ..
570812 -            [On 000420 Eric Armstrong lists DKR systems. ref SDS 79
570813 -            4877
570815 -  ..
570816 - Jack says in his second letter received this morning, per above,
570817 - ref SDS 0 6370, that he has
570818 -
570819 -      "...a great time roaming the welchco web site, but I occasionally
570820 -      have fits of discomfort with some of what I see.  I suspect that
570821 -      this discomfort provokes early comments I have on the
570822 -      stream-of-consciousness (highly linked, mind you) that I see
570823 -      there. ref DRT 1 0001
570825 -  ..
570826 - Jack describes the effort to maintain alignment as...
570827 -
570828 -     ...Rod's opinions mixed with links to other opinions, linked
570829 -     ultimately to original sources. ref DRT 1 3780
570830 -
570831 -     I have no discomfort with the opinions I see expressed, though it
570832 -     is worth noting that I do not necessarily agree with some of the
570833 -     conclusions drawn or associations made;, ref DRT 1 4608;
570834 -
570835 -            [On 000504 Jack Park complains to OHS/DKR team struggling
570836 -            to formulate a design for Knowledge Management has ignored
570837 -            SDS, and foundational documents as "things we don't talk
570838 -            about around here," and noted requirements to ponder SDS in
570839 -            order to accomplish project objectives. ref SDS 84 XD5M
570841 -             ..
570842 -            [On 000922 Jack baffled about why people have not addressed
570843 -            SDS foundational documents, POIMS, NWO, Typical Day
570844 -            Scenario, even if they do not try SDS, ref SDS 96 016R; he
570845 -            feels peer pressure may be a factor; people feel funny
570846 -            "going first" to cite authority for ideas for Knowledge
570847 -            Management that nobody else has presented. ref SDS 96 FJ5H
570849 -             ..
570850 -            [On 000907 Jack presents research showing time,
570851 -            chronlology, sequence applied with diary and journal
570852 -            methods help people track "trains of thought," ref SDS A3
570853 -            J69I; Jack cites Welch SDS records on Internet for applying
570854 -            this method. ref SDS 95 E37M
570856 -             ..
570857 -            [On 001116 Jack Park letter to OHS/DKR team meeting at SRI
570858 -            reports SDS records on the Internet demonstrate amazing
570859 -            ability to accomplish Knowledge Management that validates
570860 -            claims in POIMS. ref SDS A1 953G
570862 -  ..
570863 - Jack seems unaware that SDS records are prepared with eight (8) steps
570864 - for converting "stream-of-conscious" information into useful
570865 - knowledge, reported on 001219, ref SDS A6 4W4L, and presented in
570866 - POIMS. ref OF 3 685K
570868 -  ..
570869 - Some "discomfort" reported by Jack aids communication by spurring
570870 - feedback to get an effective record, discussed with the US Army Corps
570871 - of Engineers on 961017. ref SDS 12 5832  The challenge is to strike a
570872 - balance between traditional "feel good" management, discussed on
570873 - 911123, ref SDS 4 1331, and excessive rigidity that leads to fear of
570874 - accountability, discussed on 980405, ref SDS 18 5065, recognizing that
570875 - balance is never attained.  It is a constant quest of leadership.
570877 -  ..
570878 - Jack notes effort to improve stream-of-conscious communication cited
570879 - in the letter this morning, ref DIT 1 0870, and linked to analysis on
570880 - 990924 of medical mistakes, that sets out weakness of conventional
570881 - email, and that adding a process for regular alignment can help
570882 - augment, ref DIP 2 1045, per Drucker who proposes "routinizing" better
570883 - methods, from work on 991025. ref SDS 32 0785  Improving email is a
570884 - goal of Doug's augment efforts, per Eric reviewed on 000120.
570885 - ref SDS 37 3871
570887 -  ..
570888 - Jack seems to find the Welch record is a "stream-of-conscious
570889 - rendering that is highly linked; he does not comment on the purpose or
570890 - value added of "connecting the dots" with respect to helping maintain
570891 - alignment to objectives, requirements and the experiential record,
570892 - i.e., history.
570893 -
570894 -     [On 000419 debate about online books illustrates resistance to
570895 -     changing the paradigm from documents to Knowledge Space.
570896 -     ref SDS 78 3599
570898 -      ..
570899 -     [On 000420 near term ease reacting to email is secuctive, moving
570900 -     people to avoid the hard word of adding connections that convert
570901 -     information into "intelligence" positioned in Knowledge Space.
570902 -     ref SDS 80 4911
570904 -  ..
570905 - Jack does not set out any incorrect representations, which may
570906 - indicate that none were found, or merely that he did not have time nor
570907 - inclination to request corrections.  His letter is somewhat lengthy,
570908 - which indicates that if important errors were encountered, they could
570909 - have been identified within the time expended on his letter, leading
570910 - to an inference that no corrections were found to be necessary.
570912 -  ..
570913 - The New World Order... paper explains the purpose of linking to
570914 - original sources is to discover alignment that imparts understanding,
570915 - and verifies accuracy, which in turn builds and maintains shared
570916 - meaning, ref OF 5 7492, essential for collaboration.  On 900319 there
570917 - was a report that an accurate record improves human reasoning.
570918 - ref SDS 2 0702  More recently on 950925 Doug Hoffstadter's work showed
570919 - a similar conclusion, particularly with respect to the sequence of
570920 - events, commonly called "chronology." ref SDS 7 3888  On 000329 Doug
570921 - Engelbart commended Hoffstadter's work, indicating some confidence in
570922 - Hoffstadter's support for understanding sequence. ref SDS 67 4851
570924 -  ..
570925 - Jack's letter today seems conflicting with his letter a few days ago
570926 - on 000403 that reported SDS and POIMS are "really slick." ref SDS 72
570927 - 4862   Jack's comment that commentary in SDS records links back to
570928 - original sources shows awareness of explanations in POIMS. ref OF 3
570929 - 16EF
570930 -
570931 -            [On 000907 Jack presents research showing time,
570932 -            chronlology, sequence applied with diary and journal
570933 -            methods help people track "trains of thought," ref SDS 95
570934 -            J69I; Jack cites Welch SDS records on Internet for applying
570935 -            this method. ref SDS 95 E37M
570937 -  ..
570938 - Jack does not cite in his letter today disagreement with any opinion,
570939 - representation, nor conclusion in the Welch record.
570941 -  ..
570942 - Jack does not cite any error in tracking the sequence of events and
570943 - understandings, critical to effective reasoning.
570945 -  ..
570946 - This all sounds perhaps overly officious, but reflects the process of
570947 - "feedback" as a critical metric of communication, cited in POIMS,
570948 - ref OF 3 3774, which is essential to create an experiential record
570949 - that makes a DKR useful.  In other, words we need a few implementation
570950 - procedures and skills to guide application of technology, which can be
570951 - thought of as "Communication Metrics," for the reasons set out in
570952 - POIMS. ref OF 3 2200
570953 -
570955 -  ..
570956 - Knowledge Representation Requires Use-case Analysis
570957 - Define Opinion, Analysis, Daily Working Information, Knowledge
570958 -
570959 - Jack wants the DKR team to submit opinions on five (5) issues,
570960 - relative to the Welch method of developing "meaning" from an
570961 - experiential record, cited below in connection with the process of
570962 - "Knowledge Representation, ref SDS 0 0924, as follows....
570963 -
570964 -      [On 000420 Eric Armstrong lists DKR systems. ref SDS 79 4877
570966 -       ..
570967 -      [On 000426 Jack cites SDS as providing partial solution for DKR.
570968 -      ref SDS 82 3315
570969 -
570970 -    1.  ...how opinions are, in fact, to be represented [in DKR].
570971 -        ref DRT 1 4608
570972 -
570973 -    2.  ...[opinion] should be captured, but I suspect [opinion] cannot
570974 -        be the sole body of experience and inference represented, and
570975 -        there will have to always be more than one opinion set
570976 -        captured.
570978 -        ..
570979 -       [On 000504 Eugene explains "knowledge" is researcher's summary
570980 -       and comments on information. ref SDS 84 0007
570982 -        ..
570983 -       [...Eugene intermingles annotations with notes of original
570984 -       source material. ref SDS 84 6205
570985 -
570986 -    3.  ...welchco [is] one persons giant neural network, complete with
570987 -        narrative -- a kind of "granny cell" network. ref DRT 1 6500
570988 -
570989 -    4.  ...there may be alternatives we have not yet thought of.
570990 -
570991 -    5.  ...do we want to capture personal journals in the DKR?
570993 -  ..
570994 - Jack questions do not address SDS design for converting information
570995 - into knowledge by investing intellectual capital for personal and
570996 - organizational integrated memory and management; there is not one
570997 - reference to POIMS. ref OF 3 6221  The record on 000331 reports
570998 - discussion with Jack on SDS support for time and context. ref SDS 71
570999 - TQ68
571001 -                ..
571002 -               [On 000504 Jack noted to OHS/DKR team that SDS
571003 -               capabilities are things we don't talk about around here.
571004 -               ref SDS 84 XD5M
571006 -                ..
571007 -               [On 000922 Jack baffled about why people have not
571008 -               addressed SDS foundational documents, POIMS, NWO,
571009 -               Typical Day Scenario, even if they do not try SDS,
571010 -               ref SDS 96 016R; he feels peer pressure may be a factor;
571011 -               people feel funny "going first" to cite authority for
571012 -               ideas for Knowledge Management that nobody else has
571013 -               presented. ref SDS 96 FJ5H
571015 -            ..
571016 -           Doug Engelbart seemed to indicate on 000326 that a "journal"
571017 -           could support a DKR effort. ref SDS 62 0001
571019 -            ..
571020 -           Jack's reference on 000331 shows Geletner suggests
571021 -           chronology is a good organizing criteria for knowledge.
571022 -           ref SDS 71 5946
571024 -            ..
571025 -           "Opinion" and decision support are primary purposes of the
571026 -           DKR.  If we can "augment" opinion with alignment to original
571027 -           sources, commitments and established policy and precedent to
571028 -           privide consistency in decisions, this greatly reduces
571029 -           tension that arises from arbitrary conduct that causes
571030 -           mistakes, loss, stress and conflict.
571032 -  ..
571033 - This may relate to Jack's suggestion in his earlier letter today, see
571034 - below, proposing that the DKR team work on Knowledge Representations
571035 - for "Meaning Constructs." ref SDS 0 0924
571037 -  ..
571038 - Jack poses critical questions that help address Eric Armstrong's query
571039 - on 000120 about what a DKR looks like, ref SDS 37 3002, which requires
571040 - defining....
571041 -
571042 -        •  Information
571043 -
571044 -        •  Analysis
571046 -            ..
571047 -        •  Knowledge
571049 -            ..
571050 -        •  Opinion (decision making, ref SDS 0 7722)
571052 -            ..
571053 -        •  Journal, diary, daily working information, the "record" and
571054 -           "intellectual capital"
571056 -  ..
571057 - Are these different things, the same thing, overlapping things?
571058 -
571059 -    On 960227 definitions were developed. ref SDS 9 0022  An example is
571060 -    in the record on 900303. ref SDS 1 7739
571061 -
571062 -    How much time, expense, mental and emotional capital should be
571063 -    applied to separate "knowledge" from analysis, and opinion from
571064 -    "daily working information" cited by Doug Engelbart as part of
571065 -    knowledge management, reviewed on 000327. ref SDS 63 3971
571067 -     ..
571068 -    Below, Eric proposes use-case analysis to address this question,
571069 -    ref SDS 0 0784, per Jack's recommendation on 000324. ref SDS 59
571070 -    6052
571071 -
571072 -       [On 000420 Lee Iverson explains Use Case analysis during
571073 -       project meeting. ref SDS 80 4933
571075 -        ..
571076 -       [On 000504 Eugene Kim intermingles annotations with notes of
571077 -       original source material. ref SDS 84 6205
571078 -
571080 -  ..
571081 - Background that may help address this question...
571082 -
571083 -      POIMS explains "intelligence" converts information into knowledge
571084 -      by investing "intellectual capital. ref OF 3 6649
571085 -
571086 -      On 000328 letter to Doug Engelbart and post-Colloquium project
571087 -      team at SRI suggests hearing from cognitive science people for
571088 -      help on defining "knowledge." ref SDS 65 0001
571090 -       ..
571091 -      On 000327 Doug Engelbart pointed the team to his paper published
571092 -      in 1972 that provides a comprehensive explanation of "knowledge
571093 -      work" in relation to "daily working information. ref SDS 63 3971
571095 -       ..
571096 -      On 000326 Doug Engelbart seemed to indicate that a "journal"
571097 -      could support a DKR effort. ref SDS 62 0001
571098 -      ..
571099 -      On 000324 record of first post-Colloquium meeting at SRI of
571100 -      project team shows requirement to define DKR. ref SDS 59 3611
571102 -       ..
571103 -      On 000120 letter to Colloquium suggested defining "knowledge."
571104 -      ref SDS 37 5063
571106 -       ..
571107 -      On 931130 review of Peter Drucker's work shows "analysis" is a
571108 -      major responsibility of management. ref SDS 6 7911
571110 -       ..
571111 -      On 900303 review of a book "The Improbably Machine" explains that
571112 -      "stories" are a big part of human intelligence. ref SDS 1 3016
571113 -
571114 -
571115 -
571116 -
571117 -
571118 -
571119 -
5712 -

SUBJECTS
Diary Experience Integrate Documents
Chronology, Cause and Effect, Sequence Complexity
Clear Concise Complete Communication Judicious Review Avoid being Ove
DKR Dynamic Knowledge Repository Organizational Memory Capture Experi
SDS Unique Nobody Else Doing Knowledge Management
SDS Unique Only Example of Anyone Doing KM, Bill Bearden
SDS Secret of KM Others Don't Have It
SDS Killer Application Improves Alphabet Technology Like Alphabet Imp
New Way Thinking Through Writing Write Story to Construct Meaning Con

9211 -
921201 -  ..
921202 - Experiential Record, Like SDS, Useful for Knowledge Management
921203 -
921204 -
9213 -
9214 -
9215 - 1333 received letter
9217 -    ..
921701 - Received ref DRT 2 0001, from Bill Bearden responding to Jack's
921702 - letter, ref DRT 1 0001, and commenting...
921703 -
921704 -     DKR = [Recorded Dialog] + [Outside Intelligence] + [Knowledge
921705 -     Product]
921707 -      ..
921708 -     A journal isn't really a "dialog", I guess. But I would think my
921709 -     personal DKR would include my journal.  Perhaps it wouldn't all be
921710 -     published to the world as Rod's is. And it might not look exactly
921711 -     like Rod's does. But it seems to me that Rod is on the right track
921712 -     as far as personal information/knowledge goes. Who among us is
921713 -     doing anything as sophisticated? (a rhetorical question, not a
921714 -     challenge :-)
921716 -  ..
921717 - Bill's comments today support incorporating time as a primary design
921718 - requirement for knowledge management (i.e., KM, DKR), based on the
921719 - model of a journal.
921721 -  ..
921722 - At the project meeting on 000330 Marcello Hoffman seemed to indicate
921723 - there are no KM applications, which aligns with Bill's comment today
921724 - indicating that SDS is an advanced system.  This is more evidence that
921725 - supports Dick Karpinski's letter on 000227 reporting SDS is the best
921726 - KM program available. ref SDS 48 0897
921727 -
921728 -       [On 000425 evidence indicates that SDS is the only program that
921729 -       supports KM. ref SDS 81 0480
921731 -        ..
921732 -       [On 000504 minimal response to Jack's request for comments on
921733 -       SDS, per above, ref SDS 0 2928, led Jack to comment that SDS
921734 -       capabilities are things we don't talk about around here.
921735 -       ref SDS 84 AJ8H
921737 -  ..
921738 - Bill, however, does not directly address Jack's threshold question of
921739 - whether "opinion" should be in the DKR. ref SDS 0 2928
921741 -  ..
921742 - Assuming "opinion" should not be in the DKR, what distinguishes
921743 - opinion from other stuff, and what happens to it?  The letters
921744 - submitted through the Colloquium might be construed to contain
921745 - "opinion."  Would these all be deleted?
921747 -  ..
921748 - What would be wrong with connecting up the resource of daily opinion
921749 - as part of daily working information that offers clues for for guiding
921750 - daily action and long-term planning?
921752 -  ..
921753 - For example court opinions are saved.  If the opinion of the court is
921754 - not saved in the DKR, where should it be saved so that people can
921755 - apply that resource for future similar circumstances, and to guide
921756 - formulation of future opinion?
921757 -
921758 -
921759 -
921760 -
921761 -
921762 -
9218 -

SUBJECTS
SDS Diary Contributes to DKR Objectives Responding to Jack Park's Req
Email is Diary of Interactions Journal SDS Contributes to DKR Objecti

9804 -
980501 -  ..
980502 - Journal Incorporated into DKR because Email is a Diary of Interactions
980503 - Email is a Form of Diary that Can be Incorporated into DKR
980504 - DKR Can Incorporate Personal Journal Since Email is a Form of Diary
980505 -
980506 -
980507 -
9806 -
9807 -
9808 - 2115
9810 -    ..
981001 - Received ref DRT 5 0001 from Eric Armstrong saying...
981002 -
981003 -      There isn't any harm in the DKR capturing personal journals, but
981004 -      he is not sure how valuable it is.
981005 -
981006 -        [000414 Eric recalled feeling that "hypertext journal system"
981007 -        is not an impressive capability, but discovered from observing
981008 -        a demonstration that Traction is a "diamond." ref SDS 77 5850
981010 -         ..
981011 -        [On 000420 Eric supplements prior favorable review saying
981012 -        Traction is more of a journaling tool than for interactive
981013 -        discussion. ref SDS 79 6319
981015 -         ..
981016 -        [On 000420 Eric lists DKR systems. ref SDS 79 4877
981018 -       ..
981019 -      Eric's requirements support publishing from private DKRs into the
981020 -      public version, and notes a need to track where private material
981021 -      has been published so that only version-differences are sent when
981022 -      publishing subsequently. ref DRT 5 3360
981024 -  ..
981025 - Eric notes that...
981026 -
981027 -      ...an email forum is a "diary" of interactions.
981028 -
981029 -          On 000404, yesterday, Eric proposed tools to capture
981030 -          associated ideas easily and naturally, without having to
981031 -          "change context" to another document to do so. ref SDS 73
981032 -          4561
981034 -           ..
981035 -          Eric's recognition of an email forum as a form of diary seems
981036 -          to support using a diary or journal as part of the DKR, per
981037 -          above, ref SDS 0 3877, and aligns with Doug Engelbart letter
981038 -          received on 000326 indicating a "journal" might support a DKR
981039 -          effort. ref SDS 62 0001
981041 -           ..
981042 -          Eric does not differentiate "diary" from "opinion" and other
981043 -          forms of knowledge, per above. ref SDS 0 0448
981045 -       ..
981046 -      The reduction step consists of abstracting the thoughts they
981047 -      contain, effectively replacing them in a subsequent version of
981048 -      the document. ref DRT 5 6972
981049 -
981050 -          This sounds like the core process of SDS which continually
981051 -          aligns new information.
981052 -
981053 -
981054 -
981055 -
9811 -

SUBJECTS
SDS Difficult to Follow because Everything is Done with Lin
Inline Links Solve Problem SDS Diary Difficult to Follow Because Ever
Link Types Enable People to Decide Not to Follow Link to Original Sou
Links Boggle Mind in Knowledge Space, Eric Armstrong
Links Boggle Mind Prefer Avoid Responsibility Online Links SDS Comman
Click Links Objectionable, Takes Time Context Adds Value Command Cont

A508 -
A50901 -       ..
A50902 -      Links in SDS Harder to Follow than Inline Information
A50903 -      SDS More Difficult to Follow Using Links than with Inline
A50904 -      Inline Material Reduces Problem of Links Improves Productivity
A50905 -
A50906 - Eric notes...
A50907 -
A50908 -      Rod's diary is difficult to follow because everything is done
A50909 -      with links. ref DRT 5 1584
A50910 -
A50911 -          The letter to Jack discussed this issue. ref DIP 1 7169 sent
A50912 -          on 000403. ref SDS 72 3705
A50914 -           ..
A50915 -          [On 000504 SDS links confuse Eric. ref SDS 83 1739
A50917 -           ..
A50918 -          [On 000517 Eric confirms SDS is a poor program because links
A50919 -          are confusing; he likes printed material. ref SDS 89 9351
A50921 -           ..
A50922 -          [On 000601 Paul Fernhout says links are confusing because
A50923 -          multiple windows are open. ref SDS 91 0894
A50925 -           ..
A50926 -          [On 000824 Eric wants email that delivers reply in context.
A50927 -          ref SDS 94 G92D
A50929 -           ..
A50930 -          [On 020530 Eric reports people will use SDS to get benefits
A50931 -          of incredible memory that is not available from other
A50932 -          methods; linking everything seems less important. ref SDS B2
A50933 -          Y89F
A50935 -  ..
A50936 - Eric continues...
A50937 -
A50938 -      What you want, though, is for much of the material to be "inline"
A50939 -      instead of linked. When you publish a "reduced" version, the
A50940 -      material that was originally inline becomes relegated to a link
A50941 -      -- a link which is identified as something that points to
A50942 -      "supporting information" or "original arguments" or the like, so
A50943 -      you know you don't want to follow it unless you are particularly
A50944 -      interested in deeper information. ref DRT 5 4473
A50945 -
A50946 -            [On 000414 Eric cites "inline" links and "transclusions" as
A50947 -            limitation of Traction. ref SDS 77 0006
A50949 -             ..
A50950 -            [On 000504 Eric reported being confused by so many links in
A50951 -            SDS. ref SDS 83 1739
A50953 -             ..
A50954 -            [On 000505 Eric's editor spec may address this issue, but
A50955 -            not clear. ref SDS 85 5003
A50957 -             ..
A50958 -            [On 000517 Eric likes inline material rather than links.
A50959 -            ref SDS 89 9351
A50961 -             ..
A50962 -            [On 000811 Eric objects to clicking on links, wants inline
A50963 -            material. ref SDS 93 0001
A50965 -             ..
A50966 -            [On 000824 Eric worried that a DKR design like SDS would
A50967 -            create "link-containing" messages, suggesting he proposes
A50968 -            fewer links. ref SDS 94 FJ5H
A50970 -             ..
A50971 -            [On 001018 Eric reported linking presents big problem for
A50972 -            KM design, an architecture snag. ref SDS 98 3X9G
A50974 -             ..
A50975 -            [On 001102 Eric reported delays in releasing KM program
A50976 -            incorporating ideas proposed today for improving SDS.
A50977 -            ref SDS A2 0001
A50979 -           ..
A50980 -          Concern about being overwhelmed following links couched in a
A50981 -          proposal for greater use of "inline" material reflects worry
A50982 -          on 000125 that links in a DKR "boggle the mind." ref SDS 39
A50983 -          3867
A50985 -           ..
A50986 -          Knowledge Management Dilemma Resolved by Better Technology
A50987 -          Intelligence Mind Boggling to the Conscious Span of Attention
A50988 -
A50989 -          Objections to using Knowledge Space by clicking on
A50990 -          links present a core dilemma of Knowledge Management...
A50991 -
A50992 -              How to deliver a greater share of the connections created
A50993 -              by intelligence that grow knowledge of cause and effect
A50994 -              without overwhelming the conscious mind.  Increasing the
A50995 -              the power to understand context, which is easily managed
A50996 -              in the subconscious mind, overwhelms the conscious span
A50997 -              of attention until the right tools are used for Knowledge
A50998 -              Management. see NWO, ref OF 5 4925
A50999 -
A51000 -          ...evident from conflict between Eric's call today for fewer
A51001 -          links, and his letter on 000227 calling for "intelligence"
A51002 -          that connects current work to relevant history in order to
A51003 -          save time. ref SDS 48 1248
A51004 -
A51005 -              [On 011010 reviewed frustrations people have learning to
A51006 -              work with connections that grow intelligence. ref SDS B1
A51007 -              KS6G
A51009 -           ..
A51010 -          SDS technology over the past 15 years has proven effective
A51011 -          for creating and handling increased power from using
A51012 -          information to growing knowledge.  Eric is actually
A51013 -          describing today the inevitable tension and frustration of
A51014 -          learning to make this transition.
A51016 -           ..
A51017 -          Worry about being overwhelmed by links was addressed on
A51018 -          990419 by developing procedures for Judicious Review
A51019 -          ref SDS 21 2601
A51021 -           ..
A51022 -          Implementation of Eric's idea would increase duplication of
A51023 -          information that compounds error, delay and cost.  The SDS
A51024 -          design of Knowledge Space explained on 960620, ref SDS 11
A51025 -          3516, enables recylcing information to leverage the
A51026 -          architecture of human thought that saves time, avoids
A51027 -          confusion and reduces mistakes by creating connections that
A51028 -          maintain context using conservation of knowledge set out in
A51029 -          POIMS. ref OF 3 4662
A51031 -           ..
A51032 -          Would like to see work product and experience showing how
A51033 -          duplication of material "inline" increases understanding
A51034 -          and/or saves time and money.  How would this helps people
A51035 -          decide to click on a link to get more information, which is
A51036 -          the only reason to do so.  This is a critical aspect of
A51037 -          knowledge management.
A51038 -
A51039 -             [On 001102 Eric reported delays in releasing KM program
A51040 -             incorporating ideas proposed today for improving SDS.
A51041 -             ref SDS A2 0001
A51043 -              ..
A51044 -             [On 010916 Eric reported SDS enables amazing memory to
A51045 -             find everything with mechanisms that obviousy work.
A51046 -             ref SDS A9 0001
A51048 -              ..
A51049 -             [On 011003 Eric reported productivity is paralyzed using
A51050 -             other methods, people cannot find anything. ref SDS B0
A51051 -             EC5N
A51053 -              ..
A51054 -             [On 020530 Eric reports people will use SDS to get
A51055 -             benefits of incredible memory that is not available from
A51056 -             other methods; objectiions to linking everything seems
A51057 -             less important. ref SDS B2 Y89F
A51058 -
A51059 -
A51060 -
A511 -

SUBJECTS
SDS Difficult to Follow because Everything is Done with Lin
Inline Links Solve Problem SDS Diary Difficult to Follow Because Ever
Link Types Enable People to Decide Not to Follow Link to Original Sou
Competing Versions SDS Diaries Multiple People Doing Diaries Causes D
Use Case Analysis Eric Will Analyse Problems Caused by Multiple Peopl
SDS Multiple People Preparing Competing Diaries Diary Messy Causes Di
Conflicts Overwhelming Many People Using SDS
Links Invisible Implicit

AE10 -
AE1101 -       ..
AE1102 -      Competing Versions of Record are Messy When Many Use SDS
AE1103 -      Life Full of Competing Versions SDS Enables Early Discovery
AE1104 -      Discovery Conflict Goal of Intelligence to Avoid Mistakes
AE1105 -
AE1106 -      Eric advises that things get messy when competing versions
AE1107 -      of the summary are created by different people using "reduction."
AE1108 -      ref DRT 5 1560   He plans to tackle the problem with use-case
AE1109 -      analysis. ref DRT 5 7275
AE1111 -           ..
AE1112 -          An example of "competing versions" would be helpful, showing
AE1113 -          how this is different from people recalling the same event
AE1114 -          differently in a meeting, phone call or correspondence, like
AE1115 -          email, report, book, etc.
AE1117 -           ..
AE1118 -          Eric's concern about competing versions of SDS records was
AE1119 -          raised by Morris on 951103. ref SDS 8 9066
AE1121 -           ..
AE1122 -          Later on 961017 the problem of people having different
AE1123 -          interpretations and recollections of meaning and what was
AE1124 -          said was reviewed by US Army Corps of Engineers, as a major
AE1125 -          requirement for a method to discover competing versions of
AE1126 -          events early so that effort can be made to build and maintain
AE1127 -          shared meaning. ref SDS 12 1381
AE1129 -           ..
AE1130 -          Feedback showing competing versions are a benefit that enable
AE1131 -          early discovery of conflicting views for resolution to avoid
AE1132 -          mistakes, as explained in POIMS. ref OF 3 IE6L
AE1134 -              ..
AE1135 -             [On 000407 listed agenda issue for project meetings.
AE1136 -             ref SDS 76 4757
AE1138 -              ..
AE1139 -             [On 000420 Lee Iverson explains Use Case analysis during
AE1140 -             project meeting. ref SDS 80 4933
AE1142 -              ..
AE1143 -             [On 001017 Eric developed technology that solves problem
AE1144 -             of competing versions. ref SDS 97 TP6U
AE1146 -              ..
AE1147 -             [On 001018 Eric reported versioning presents big problem
AE1148 -             for KM design, another architecture snag. ref SDS 98 V3Y1
AE1150 -              ..
AE1151 -             [On 010725 Wayne Wetzel reported advantages of SDS for
AE1152 -             discovering conflicts early to avoid mistakes. ref SDS A8
AE1153 -             KM7G
AE1154 -
AE1155 -
AE1156 -
AE1157 -
AE1158 -
AE1159 -
AE1160 -
AE12 -

SUBJECTS
Knowledge Representation, 000405
Park Recommends SDS for Meaning to Develop Knowledge Representation
Analysis Mathamatics Knowledge Representation
Analysis Language Knowledge Representation
Systems Dynamics Knowledge Representation
Meaning Constructs Knowledge Representation
Evaluate Market Conditions; Progress Prospects
Park, Reports POIMS Supports KM
SDS Develops Meaning by Managing Experience
Semiotics Knowledge is Experience Peirce, 000515
Knowledge Representation Analytical Forms System Dynamics Meaning Con

B313 -
B31401 -  ..
B31402 - Knowledge Mandala Provides Five (5) Major Components of Knowledge
B31403 - Nancy Glock-Grueneich Formulates Parts of Knowledge Representation
B31404 - Welch Formulates Process for Applying "Meaning" Component Knowledge
B31405 -
B31406 -
B315 -
B316 - 0811
B317 -
B31701 - Received ref DRT 6 0001 from Jack Park sent at 0750 suggesting vu
B31702 - graphs built by Dr. Nancy Glock-Grueneich at...
B31703 -
B31704 -         http://www.higheredge.org/mandala1.ppt
B31705 -
B31706 - ...show the big picture of knowledge representation, ref DRT 6 3294,
B31707 - she calls it a...
B31709 -                        ..
B31710 -                       Knowledge Mandala
B31711 -
B31712 - ...to depict 5 aspects of knowledge...
B31714 -         ..
B31715 -    1.  Analytical Forms Systems Dynamics Meaning Constructs Cases
B31716 -        Applications
B31718 -  ..
B31719 - This helps address Doug Engelbart's request on 000307 for an
B31720 - explanation of "knowledge," similar to Bellinger's work. ref SDS 51
B31721 - 4820  This work lends general support for explanation of "knowledge"
B31722 - in POIMS that calls for adding "intelligence" to information, which
B31723 - includes "analysis."  Below, SDS support for case studies reflects
B31724 - "experience" as a component of "knowledge." ref SDS 0 P75M
B31725 -
B31727 -     ..
B31728 -    Ontology Needed for Analytical Forms, Systems Dynamics
B31729 -
B31730 -    Jack says that the first two (2) parts...
B31731 -
B31732 -                •  Analytical Forms
B31733 -                •  Systems Dynamics
B31734 -
B31735 -    ...concern ways of representing things using signs, symbols,
B31736 -    ontologies, and so forth.  This level requires a uniform ontology.
B31737 -    ref DRT 6 1722
B31739 -     ..
B31740 -    Need example of an ontology that uses "signs" and "symbols."
B31741 -    Possibly he is thinking of mathamatical operators, e.g., + - x
B31742 -
B31743 -       [On 000515 Peirce's semotic theory of philosopny uses ontology
B31744 -       to examin the structure of the world (elements and relations of
B31745 -       existence. ref SDS 87 0784
B31747 -        ..
B31748 -       [On 000709 Cliff Joslyn cited semiotics encompass signs and
B31749 -       synmbols for inference, meaning and communication. ref SDS 92
B31750 -       4078
B31752 -        ..
B31753 -       [On 001130 Jack proposes ontology "engine" to improve SDS
B31754 -       subject management and linking. ref SDS A4 GK8O
B31756 -        ..
B31757 -       [On 001214 Jack proposes uses cases to develop ontology under
B31758 -       Qualitative Process Theory (QPT). ref SDS A5 0001
B31760 -     ..
B31761 -    On 000221 Jack cited "Ontology" as a requirement for Knowledge
B31762 -    Management, ref SDS 46 8044, which is defined from work in AI as a
B31763 -    "common vocabulary." ref SDS 46 2622, and proposed an "interlingua"
B31764 -    process. ref SDS 46 4350  He explained that creating ontologies is
B31765 -    difficult, like "Pandora's Box, ref SDS 46 7455, reflecting POIMS
B31766 -    explanation of human "intelligence." ref OF 3 0561  Below, Paul
B31767 -    Faunhout concurs that Knowledge Representation is a very difficult
B31768 -    task. ref SDS 0 0005
B31770 -     ..
B31771 -    On 000227 Eric Armstrong asked how to organize "case study"
B31772 -    material and other sources in, books, magazines, journals, papers,
B31773 -    and Web articles, and expects this would improve daily work enough
B31774 -    to establish eventual supremacy of knowledge management in a
B31775 -    "survival of the fittest" business climate. ref SDS 48 0937
B31776 -
B31778 -     ..
B31779 -    Meaning From Experience Supported by Welch's SDS Program
B31780 -    Case Studies Derivitive of SDS May Support KR "Cases"
B31781 -
B31782 -    Follow up ref SDS 71 TQ68.
B31783 -
B31784 -    Jack says knowledge representation makes "meaning" (semantics) out
B31785 -    of experience, and cites the Welch...
B31786 -
B31787 -           http://www.welchco.com
B31788 -
B31789 -    ...system for "enumeration of experience." ref DRT 6 1517  Jack
B31790 -    sent a second letter today asking the project team to assess the
B31791 -    role of opinion and a journal for knowledge management, based on
B31792 -    the Welch system. ref SDS 0 2928
B31794 -     ..
B31795 -    This supplements the report on 000227 citing SDS for effective
B31796 -    knowledge management. ref SDS 48 XG6J and on managing high volume
B31797 -    of correspondence. ref SDS 48 XH71
B31798 -
B31800 -     ..
B31801 -    Use Cases for Analytic, System, Meaning Knowledge Representation
B31802 -
B31803 -    Jack proposes that the DKR team develop Knowledge Representation to
B31804 -    enumerate use cases for...
B31805 -
B31806 -        •  Analytical Forms,
B31807 -        •  Systems Dynamics, and
B31808 -        •  Meaning Constructs.
B31809 -
B31810 -    ...and that this be the topic #2 for of Knowledge Representation.
B31811 -    ref DRT 6 1728
B31813 -       ..
B31814 -      Jack may have launched this discussion by raising the Welch
B31815 -      "journaling" and "opinion" system cited in his second letter
B31816 -      today, reviewed above. ref SDS 0 2928
B31818 -       ..
B31819 -      A possible example of "cases" might be "case studies" and "use
B31820 -      cases" which accumulate experience based on subject context.  The
B31821 -      U.S.  Army Corps of Engineers published a report that
B31822 -      Communication Metrics supported by SDS, improves "Lessons
B31823 -      Learned," which reflects the value added of "case studies."
B31824 -      ref DRP 2 3156
B31826 -     ..
B31827 -    Order of Discussion Knowledge Representation
B31828 -
B31829 -    Jack proposes...
B31830 -
B31831 -        Analytic Forms = Knowledge Representation 3
B31832 -        Systems Dynamics = Knowledge Representation 4
B31833 -        Meaning Constructs = Knowledge Representation 5
B31834 -        Applications-DKF = Knowledge Representation 6
B31835 -        Applications-OHS = Knowledge Representation 7
B31836 -        Serialization (XML, etc) = Knowledge Representation 8
B31837 -
B31838 -
B319 -
B31901 -
B320 -

SUBJECTS
XML Limitations, Xanadu, Nelson
XML Helpful, No Easy Solution Knowledge Repesentation, Paul Fernhout
Knowledge Repesentation Difficult, Paul Fernhout
Knowledge Representation Difficult Human Intelligence Complex
Knowledge Representation Not Solved by XML

BE07 -
BE08 - 1548
BE0901 -  ..
BE0902 - XML Helpful, But No Easy Solution for KM Design
BE0903 - Knowledge Representation Difficult Problem, No Industry Answer Yet
BE0904 -
BE0905 - Received ref DRT 3 0001 from Paul Fernhout addressed to the
BE0906 - Colloquium meeting now at SRI.  Paul seems to be responding to Eric
BE0907 - Armstrong about using XML for creating the editor and defining...
BE0908 -
BE0910 -                       ..
BE0911 -                      knowledge representation
BE0912 -
BE0913 -
BE0914 - ...which is requested by Jack Park, per above. ref SDS 0 2928
BE0916 -  ..
BE0917 - Paul's key points are...
BE0918 -
BE0919 -     Knowledge Representation is a sort of AI, even if it is designed
BE0920 -     for primarily augmenting rather than acting independently.
BE0921 -     ref DRT 3 1748
BE0922 -
BE0923 -         Need definition of "knowledge" per questions on SDS...
BE0924 -         ref SDS 0 0448
BE0926 -          ..
BE0927 -         How is "knowledge" and "managing" knowledge different from the
BE0928 -         data base program, and the information management program?
BE0930 -          ..
BE0931 -         These are key issues.....
BE0933 -      ..
BE0934 -     Knowledge management/representation problem is a deep one, and XML
BE0935 -     doesn't address it in a serious way, and confuses the subject by
BE0936 -     the hype making it sound like XML does address the topic of
BE0937 -     knowledge representation in a serious way.  However, sorry, as an
BE0938 -     alternative (not to XML, but to lots of hard work mucking in these
BE0939 -     K.R. fields) I have no easy answer. ref DRT 3 4158
BE0941 -      ..
BE0942 -     Paul's report today aligns with the record on 000305 reviewing
BE0943 -     Xanadu that indicate's Ted Nelson believes XML is not adequate for
BE0944 -     KM. ref SDS 49 7834   On 911121 Morris described an AI system for
BE0945 -     engineering management that was not useful, which he proposed
BE0946 -     could accomplish SDS support. ref SDS 3 8809
BE0947 -
BE0948 -           [On 000425 KM secret of SDS. ref SDS 81 0480
BE0950 -            ..
BE0951 -           [On 000513 sent letter to project citing Paul's comments
BE0952 -           today. ref SDS 86 0001
BE0954 -            ..
BE0955 -           [On 000513 Paul responded favorably. ref SDS 88 0001
BE0957 -            ..
BE0958 -           [On 000524 Paul submits code for information storage
BE0959 -           architecture, evidently called Pointrel Data Repository
BE0960 -           system. ref SDS 90 0783
BE0962 -            ..
BE0963 -           [On 000601 Paul comments on analysis of this record, and
BE0964 -           reference to Pointrel reviewed on 000524. ref SDS 91 0851
BE0966 -            ..
BE0967 -           [On 001025 Paul explains that creating useful tools for
BE0968 -           knowledge work is difficult because human thought and
BE0969 -           experience is complex. ref SDS 99 KY5K
BE0971 -            ..
BE0972 -           [On 001027 cites Welch work on KM and prospect that poetry
BE0973 -           may contribute to developing useful tools. ref SDS A0 0001
BE0975 -            ..
BE0976 -           [On 001124 Ken Holman concurs XML not solution for KM.
BE0977 -           ref SDS A3 0001
BE0979 -            ..
BE0980 -           [On 010131 Cliff Joslyn notes PDF and wordprocessing do not
BE0981 -           support KM. ref SDS A7 LF7M
BE0983 -         ..
BE0984 -        On 000221 Jack Park advised that organizing knowledge, what he
BE0985 -        calls "ontology" is a Pandora's Box of complexity, ref SDS 46
BE0986 -        7455, also cited above. ref SDS 0 1776
BE0988 -      ..
BE0989 -     Tools to build with, ref DRT 3 3008, needs to be based on...
BE0990 -
BE0991 -        • convenience
BE0992 -        • cross-platform
BE0993 -        • open-source
BE0994 -        • open-ended
BE0995 -        • active development communities
BE0997 -      ..
BE0998 -     There are person-millennium of AI and cognitive science work years
BE0999 -     at my feet and I can't really point to much of it as being
BE1000 -     immediately useful.  In general, much of it is generally useful
BE1001 -     for understanding the difficulty of the problem.  I'd think
BE1002 -     Hoffstadter's work on Fluid Concepts and Creative Analogies is a
BE1003 -     great source of ideas for a good example of the better stuff.
BE1004 -     ref DRT 3 0936
BE1005 -
BE1006 -        This aligns with Marcello Hoffman's comment at the meeting on
BE1007 -        000330 that no KM solution has been developed or even defined.
BE1008 -        ref SDS 69 0713
BE1010 -         ..
BE1011 -        On 000221 Jack Park advised that organizing knowledge, what he
BE1012 -        calls "ontology" is a Pandora's Box of complexity, ref SDS 46
BE1013 -        7455, also cited above. ref SDS 0 1776
BE1014 -
BE1016 -      ..
BE1017 -     Representations Change Over Time Hard to Program Flexibility
BE1018 -
BE1019 -     DKR/OHS needs flexibility to support representations changing
BE1020 -     over time as needs change. AI and Knowledge Management spend years
BE1021 -     studying and working to achieve this; but, solutions are
BE1022 -     incomplete. There is no easy answer yet (if ever). ref DRT 3 1377
BE1024 -      ..
BE1025 -     Multiple representational methods need to support human mind that
BE1026 -     has five or six strategies (ex. geometrical reasoning, formal
BE1027 -     logic, heuristic rules of thumb, pattern recognition, semantic
BE1028 -     networks, others). ref DRT 3 2015
BE1029 -
BE1030 -         What evidence shows the mind picks the "best" strategy, rather
BE1031 -         than the most recent one that works well enough within the
BE1032 -         time available for consideration?
BE1034 -      ..
BE1035 -     Restrictions on knowledge representation become obsolete over
BE1036 -     time.  Representation needs to adapt to user needs.  XML could be
BE1037 -     a part of a solution, but important issues go beyond that -- to
BE1038 -     standards creation and revision and communication, and to coin a
BE1039 -     phrase "data upgrading". ref DRT 3 3380
BE1041 -      ..
BE1042 -     Rather than focus on limiting representations (DTDs) we need ways
BE1043 -     to transform, extend, and simplify representations as needed (sort
BE1044 -     of along the multi-level approach I mentioned earlier).
BE1045 -     ref DRT 3 2520
BE1047 -      ..
BE1048 -     Paul proposes getting an overview of AI and knowledge management
BE1049 -     fields and how each area or major problem/topic would affect a
BE1050 -     DKR/OHS. ref DRT 3 4345
BE1052 -          ..
BE1053 -         Who can give this overview, i.e., who has figured out what
BE1054 -         needs to be solved and come forward with something that can be
BE1055 -         built and moves the ball forward?
BE1057 -          ..
BE1058 -         Marcello Hoffman considered this issue on 000330. ref SDS 69
BE1059 -         0713
BE1061 -      ..
BE1062 -     OHS/DKR project will evolve useful code to manipulate data
BE1063 -     structures that are related to knowledge representation.  We might
BE1064 -     also wish to have a survey of such existing code. ref DRT 3 1855
BE1065 -
BE1066 -         We have not yet defined knowledge representation, per above.
BE1067 -         ref SDS 0 4611
BE1069 -          ..
BE1070 -         How can there be code to manipulate something that does not
BE1071 -         yet exist?
BE1072 -
BE1074 -      ..
BE1075 -     Need Strong Solution to Easy Problem That Improves Productivity
BE1076 -
BE1077 -     Paul Suggests...
BE1079 -      ..
BE1080 -     Pick some problems which can be addressed relatively easily, but
BE1081 -     ideally addressable in such a way as the approach can later be
BE1082 -     revised and expanded. ref DRT 3 5226
BE1083 -
BE1084 -         This is a strong point.
BE1085 -
BE1086 -         The record for the meeting on 000324 suggested finding
BE1087 -         something that is easy to accomplish and makes a big impact on
BE1088 -         productivity. ref SDS 59 3498
BE1089 -
BE1091 -  ..
BE1092 - Knowledge Representation Classes Enumerated
BE1093 -
BE1094 -
BE11 -
BE12 -
BE13 - 1921
BE15 -    ..
BE1501 - Received ref DRT 4 0001 from Paul Fernhout responding to a letter
BE1502 - from Jack Park suggesting that the DKR development team....
BE1503 -
BE1504 -      ...enumerate the problems of knowledge representation we want
BE1505 -      to solve, then start from there.
BE1507 -         ..
BE1508 -        [On 000601 Paul encountered this record in connection with
BE1509 -        reviewing analysis of his letter on 000524 explaining Pointrel.
BE1510 -        ref SDS 91 0851
BE1512 -  ..
BE1513 - Paul sets out following...
BE1514 -
BE1515 -    1.  Entities (quantity, change, categories, existence)
BE1516 -
BE1517 -        Ex.  You're reading a mystery novel (building up knowledge
BE1518 -        structures in a DKR as you go along) and as the end it says
BE1519 -        "The butler did it".  What happens to those structures?
BE1520 -        ref DRT 4 1989
BE1521 -
BE1522 -           I don't understand how Paul's example exemplifies "entities"
BE1524 -            ..
BE1525 -           Paul might provide an example of a "knowledge structure"
BE1526 -           that one would build from reading a novel.
BE1528 -            ..
BE1529 -           On 000223 Henry van Eykan asked how to organize this type of
BE1530 -           information. ref SDS 47 1254
BE1532 -            ..
BE1533 -           Another example is a little easier, reading a project spec
BE1534 -           or a letter from the Colloquium on the DKR.
BE1536 -            ..
BE1537 -           Typically, information is organized to facilitate action,
BE1538 -           within an organic structure of needs, and objectives, per
BE1539 -           POIMS. ref OF 3 0561  Subjects and relationships created
BE1540 -           while reading comprise "intellectual capital" invested for
BE1541 -           use in performing objectives, and to build more structures as
BE1542 -           new experience dictates.
BE1544 -            ..
BE1545 -           So the answer is that the new knowledge is used to get the
BE1546 -           work done and grow more knowledge for getting more work done
BE1547 -           later, within the plan, perform, report management cycle.
BE1549 -         ..
BE1550 -    2.  Separating symbols and things
BE1551 -
BE1552 -        Ex.  When I say "tomato" that isn't everything I know about
BE1553 -        Tomatoes or everything the DKR knows about them....  Somehow
BE1554 -        that symbol can access additional information.  How is this
BE1555 -        relationship between symbol and related data maintained? (And
BE1556 -        of course, we have to acknowledge, you can't eat the symbol
BE1557 -        "tomato".)
BE1559 -            ..
BE1560 -           Let's not strive for a perfect solution, but to augment by
BE1561 -           providing a variety paths for identifying relationships and
BE1562 -           retrieving them.
BE1564 -         ..
BE1565 -    3.  Naming (uniqueness, scope, qualifiers)
BE1566 -
BE1567 -        Ex.  You say "tomato" and refer to a vegetable, and enter a
BE1568 -        reference into the DKR, I say "tomato" and refer to a set of
BE1569 -        nutrition information and try to enter that reference in the
BE1570 -        same DKR.  How do we keep these same named references referring
BE1571 -        to different things?  And likewise, if I use the scientific
BE1572 -        name for Tomato (?) how do we have it point to the same thing
BE1573 -        as some other reference to "tomato".
BE1574 -
BE1575 -           Provide links to context that clarifies distinctions. see
BE1576 -           also. ref SDS 0 1050
BE1578 -         ..
BE1579 -    4.  Representing relationships (items, domain, role, issues if
BE1580 -        computed)
BE1582 -         ..
BE1583 -        Ex. You make a link to the Tomato nutritional information from
BE1584 -        a report in the DKR.  How is this relationship maintained?
BE1585 -        Some of the vegetables you reference have pictures, how do you
BE1586 -        build this link (relationship) when you didn't plan for it?
BE1587 -        Some of the nutritional information about a tomato is
BE1588 -        calculated -- how do you distinguish this from other
BE1589 -        non-computed information?
BE1590 -
BE1591 -           Do not have a good feel for this problem.  Once a link is
BE1592 -           created it is hardwired for 10 minutes, 10 days, or 10
BE1593 -           years.  For pictures, can calculations, etc., identify the
BE1594 -           location in Knowledge Space, describe the picture, where it
BE1595 -           came from, and why it is important.  Gather evidence that
BE1596 -           supports understanding.  Move on.
BE1597 -        ..
BE1598 -    5.  Attributes to objects vs. relationships
BE1599 -
BE1600 -        Ex.  Tomato's have weight.  If you are inventorying your tomato
BE1601 -        harvest in a DKR, should you have a tomato class which has a
BE1602 -        weight attribute, or should you model this had a tomato being
BE1603 -        in a relationship with a weight?
BE1604 -
BE1605 -           This is beginning to get into DBMS.  Must the DKR manipulate
BE1606 -           data.  There are a lot of ways to do that.  We want to
BE1607 -           augment human reasoning, not DBMS. see above. ref SDS 0 1050
BE1609 -         ..
BE1610 -    6.  Why data and its description can't really be separated
BE1611 -
BE1612 -        Ex. You want to send someone your tomato nutritional
BE1613 -        information so they can store it in another DKR.  But in order
BE1614 -        to use it, they need to know about what format you stored it
BE1615 -        in.
BE1617 -            ..
BE1618 -           Write a letter explaining it.  Doesn't DLL help on this???
BE1619 -           see above. ref SDS 0 1050
BE1621 -         ..
BE1622 -    7.  Handling types and categories and sets
BE1623 -
BE1624 -        Ex. (Weak!) What does it mean that nutritional information is
BE1625 -        about a type of Tomato?  Does that then mean nutritional
BE1626 -        information can't be a type of empirical observation?  Can it
BE1627 -        belong to the set of things your friend Fred entered for you,
BE1628 -        while ate the same time being in the category of things Jim
BE1629 -        fact checked?  How is all this represented?
BE1631 -         ..
BE1632 -        Are categories just relationships in disguise?
BE1633 -
BE1634 -           Fred and Jim write up what they did, when they did it, what
BE1635 -           they hoped to achieve, what problems they encountered, who
BE1636 -           helped and any other material information.  They provide
BE1637 -           links showing alignment with relevant background and
BE1638 -           objectives for context.  This will improve productivity by
BE1639 -           an order of magnitude. see above. ref SDS 0 1050
BE1641 -         ..
BE1642 -    8.  Modeling -- both modeling reality and modeling data
BE1643 -
BE1644 -        Ex.  Simulating the movement of air over a wing vs. discussing
BE1645 -        the way you implemented this, and also realizing your
BE1646 -        simulation is not reality (the map is not the territory -- but
BE1647 -        it is its own territory).
BE1648 -
BE1649 -           Paul needs to flesh out this example a bit.
BE1651 -         ..
BE1652 -    9.  Handling elementary data (example: images or integers)
BE1653 -
BE1654 -        Ex.  Should an image stored in the DKR be made of relationships
BE1655 -        of bytes (and those relationships of bits?) or should it just
BE1656 -        be an array of bytes on a hard disk.  In either case, how do
BE1657 -        you know how to interpret those bytes back into an image?
BE1659 -            ..
BE1660 -           Try both and see what works.  If neither works try something
BE1661 -           else.  This problem has not impacted use of intelligence a
BE1662 -           great deal the past 15 years.
BE1664 -         ..
BE1665 -   10.  Data integrity
BE1666 -
BE1667 -        Ex.  For a DKR on tropical medicine, how do you make sure the
BE1668 -        weights of tomato plant parts entered do not add up to more
BE1669 -        than the weight of the entire tomato plant?
BE1670 -
BE1671 -           Link the weight to whatever spec is available, so people
BE1672 -           have a chance to recognize lack of alignment.  Sounds like
BE1673 -           the problem NASA's space thing encountered that crashed on
BE1674 -           Mars, reported 991001. ref SDS 30 0001
BE1676 -         ..
BE1677 -   11.  Data processes (triggers, integrity constrains, search)
BE1679 -         ..
BE1680 -        Ex.  When you enter a record for tomato RDA by weight into your
BE1681 -        DKR, how can the DKR  be set up to automatically send email to
BE1682 -        Eric to let him know to look it over?
BE1683 -
BE1684 -           Automatic sending of stuff typically causes information
BE1685 -           overload rather than hoped for adequate scrutiny. see above.
BE1686 -           ref SDS 0 1050
BE1688 -         ..
BE1689 -   12.  Handling multiple versions of things or patterns
BE1690 -
BE1691 -        Over time, we are going to get better information about tomato
BE1692 -        nutritional values.  How do we keep all these versions of
BE1693 -        information?  Which one do I give you when you ask for
BE1694 -        nutritional information on the tomato?
BE1696 -            ..
BE1697 -           Typically this depends upon context.  A thorough record
BE1698 -           segmented into chunks that focus attention on narrow issues
BE1699 -           at hand, provides the best opportunity to understand context
BE1700 -           for making a good decision.
BE1702 -         ..
BE1703 -   13.  Testing for sameness (equality)
BE1704 -
BE1705 -        How do I know that this reference to tomato nutritional
BE1706 -        information points to the same "thing" that this reference to
BE1707 -        "red fruit" nutritional information points to?
BE1708 -
BE1709 -           Click on the link and take a look.
BE1711 -         ..
BE1712 -   14.  Efficiency
BE1713 -
BE1714 -        Ex.  Same as handling elementary data, also how does how the
BE1715 -        system looks to a user differ from how it is implemented.
BE1716 -
BE1717 -           Critical question for Eric.
BE1718 -
BE1719 -
BE1720 -
BE1721 -
BE1722 -
BE1723 -
BE1724 -
BE1725 -
BE1726 -
BE18 -
BE19 -