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: September 7, 2006 01:26 PM Thursday; Rod Welch

Gary presents automated subject indexing for natural structures.

1...Summary/Objective
2...Jack Park Expert Organize Complexity Topic Maps Research Publications
3...Joslyn Presentation to OHS/DKR at SRI on 000727
4...Complexity Theory Presented at SRI for OHS/DKR
5...Purpose Organize Record Find Details Accuraacy Understand Cause Effect
6...New Way Working Intelligently Subject Indexing Needs Examples
7...Subject Indexing New Way Working Customers Need Examples
8...Customers Examples Work Product Functionality Learn New Way Working
9...Federating Complex Entailments Needs Explanation Examples
10...Experience Doing Knowledge Management Yieds Ideas on Federating SDS
11...Subject Indexing Cross-references Outlining Hierarchies
12...Email Subject Indexing Natural Catagories Context Management Easier
13...Integrating Information Streams Makes Better Sense of Email
14...Complementarity Organizing Record Makes Sense for Authors and Readers
15...Subject Indexing Natural Structures Organize Complexity Daily Work
16...Natural Structures Make Sense Complexity Integration Complementarity
17...Automatic Organization Email Helps Make Sense Complex Communications
18...Knowledge Management Hard Work Construct Assign Subjects
19...Metadata Assign Subjects Facilitate Finding Retrieval Hard Work
20...Sensemaking Subject Indexing Context Management Organic Structure
21...Automatic Subject Indexing Document Content Complex Context Evolves
22...Big Projects Degrade Management to Enropy Technology Cannot Keep Up
23...People Giving Up Using Technology for Communication Too Complex
24...Records Management Augments Intelligence High Volume Overwhelming
25...SDS Utopia Everything in Right Place at Right Time
26...Pain Using Management Technology Ignore Bumbling Losing Time Money
....Technology Augment Intelligence Records Management
....History Records Management Bush Proposed Trails Associations
....Roman Empire History Records Management
27...Interoperability Common Storage Enable Knowledge Management
28...Meta Data Index Finding Information Requires Common Storage
29...Common Storage Interoperability Complement Subject Index Meta Data
30...Case Study Explicit Links Granular Addressability Knowledge Management
31...Granular Addressability Explicit Links Enable Precision Access
32...Connections Complement Chronology Context with Granular Addressability
33...Subject Indexing Requirements Summarized
34...Critical Mass Tools Hard to Find to Begin Organizing the Work
35...Email Incremental Step to Knowledge Management Leverage Intelligence
36...Joslyn Explains Data Structures for Knowledge Hierarchies Networks
37...Context Organic Structures Hierarchies Trees DAGS Latices Outlines
....Method and system to compare data objects

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

1...Jack's seminar on 060523 presented ideas on merging ontologies being
2...Need examples applying data structure alternatives presented in

CONTACTS 
0204 - Los Alamos National Laboratory                                                                                                                                     O-00000755 0602
020401 - Mr. Cliff Joslyn                                                                                                                                                 O-00000755 0602
020402 -
020404 - Computer Research Group (CIC-3)

SUBJECTS
Expert Subject Indexing Playing Around Too Modest Praise Expertise J

2003 -
2003 -    ..
2004 - Summary/Objective
2005 -
200501 - Follow up ref SDS B7 0000. ref SDS B5 0000.
200502 -
200503 - Discussion continued today on subject indexing arising from question
200504 - on whether to include Rod's records for new users to experience
200505 - program tools and application. ref SDS 0 QM9I  Letter to the team
200506 - commended Jack's work on topic maps that relates to discussion of SDS
200507 - subject indexing, ref SDS 0 QM9I, including Nexist and Open Iris.
200508 - ref SDS 0 K45X  Copy to Cliff at LANL asks for support on Jack's
200509 - recollection that hypergraphs and hetroarchies relate to data
200510 - structures for subject indexing. ref SDS 0 K46T  Cliff responded that
200511 - hierarchies and networks are effective data structures in this
200512 - context. ref SDS 0 557P  Everyone feels context management is hard
200513 - work; SDS makes the 8-steps for Communication Metrics faster and
200514 - easier. ref SDS 0 PE5T  Examples and experience testing are needed to
200515 - assess advantages of incorporating new ideas for SDS. ref SDS 0 I26M
200516 - Gary Johnson contributed more analysis that hierarchies are inadequate
200517 - data structures for complex records. ref SDS 0 9J3N  He explained
200518 - advantages of outlining. ref SDS 0 IA8U
200519 -
200520 -
200521 -
200522 -
200523 -
200524 -
200526 -  ..
2006 -
2007 -
2008 - Progress
2009 -
200901 - Jack Park Expert Organize Complexity Topic Maps Research Publications
200902 -
200903 - Follow up ref SDS B7 QM9I.
200904 -
200905 - On 060831 asked Gary for comments on providing Rod's records to new
200906 - users for guidance on using SDS subject index system. ref SDS B5 VP3W
200907 - On 060905 received helpful feedback from Morris, ref SDS B7 855H, from
200908 - Gary Johnson, ref SDS B7 DJ9J, and from Jack Park. ref SDS B7 QM9I
200910 -  ..
200911 - Submitted ref DIT 1 0001 responding to Jack's letter received on
200912 - 060905, ref DRP 5 0001, and reviewed in the record at that time.
200913 - ref SDS B7 QM9I  Copies were submitted to Morris, Gary, and to Cliff
200914 - Joslyn at LANL. ref DIT 1 00W3  This letter supplements the first
200915 - letter responding to Morris and Gary on 060905. ref SDS B7 M39L
200917 -  ..
200918 - Letter to Jack, and the guys says...
200919 -
200920 -    1.  "Playing around"?  Here, here! ref DIT 1 0001
200922 -         ..
200923 -    2.  Jack has written a book or two on topic maps, and contributes
200924 -        generously to professional conferences. ref DIT 1 HE5N  On
200925 -        000829 he related new ideas on organizing the record (they were
200926 -        using "categories" at that time) presented during a Knowledge
200927 -        Management conference in Montreal. ref SDS 55 G69G
200929 -  ..
200930 - This letter commends Jack's work and commitment to advance knowledge
200931 - management, following up his letter on 060905. ref SDS B7 QM9I
200933 -  ..
200934 - Attrubution was listed as a task of knowledge management in a
200935 - presentation on 000227 by Jim Spohrer representing IBM for the
200936 - Colloquium at Stanford. ref SDS 40 0987
200938 -  ..
200939 - Letter to Jack and the guys continues...
200940 -
200941 -    3.  Nexist and Open Iris demonstrate Jack's continuing evolution of
200942 -        tools to support ontology, ref DIT 1 MM6L, reviewed on 060211.
200943 -        ref SDS A7 6Z5W
200945 -  ..
200946 - Jack's work with Nexist is illustrated by the record on 030314.
200947 - ref SDS 84 5560
200948 -
200949 -            [...see below Gary Johnson cites Jack's work with Open Iris
200950 -            applying multiple hierarchies. ref SDS 0 IA8Y
200951 -
200952 -
200953 -
200954 -
2010 -

SUBJECTS
Complexity Theory Presented But Not Heterarchies Hypergraph Presente

3103 -
310401 -  ..
310402 - Joslyn Presentation to OHS/DKR at SRI on 000727
310403 - Complexity Theory Presented at SRI for OHS/DKR
310404 -
310405 - Follow up ref SDS B7 W955.
310406 -
310407 - Letter to Jack and the guys continues...
310409 -         ..
310410 -    4.  Cliff Joslyn gave a presentation at SRI to Doug's ad hoc team
310411 -        on 000727.  Looks like Jack was the leader bringing this about.
310412 -        ref SDS 52 5555, ref DIT 1 OH7H
310414 -         ..
310415 -    5.  The record does not mention "hypergraph," but does indicate
310416 -        this might have been presented, since Cliff was clearly
310417 -        proposing ideas on complexity theory, ref SDS 52 6992, as Jack
310418 -        recalls in his letter the other day on 060905. ref SDS B7 W955
310419 -        Cliff focused his remarks to the SRI folks on chronology, which
310420 -        models the irreversibility of time that drives the locality
310421 -        principle underlying the SDS design for converting information
310422 -        into the "power of knowledge," reviewed well after Cliff's
310423 -        presentation, and on 040312. ref SDS 94 YH4G, ref DIT 1 LE7N
310425 -         ..
310426 -    6.  By copy we can ask Cliff to clarify, if he has time to sift
310427 -        through his notes.  As well, Cliff may have come back the next
310428 -        year in 2001 and discussed hypergraph, as Jack now remembers.
310429 -        I only show the one trip, but was not following OHS/DKR
310430 -        activity after 2000. ref DIT 1 LF9J
310431 -
310432 -            [...below, Cliff does not address Jack's recollection of
310433 -            presenting hypergraphs. ref SDS 0 557P
310435 -             ..
310436 -            [...below, Cliff supported discussion today citing methods
310437 -            for hierarchy data objects modeled as DAGs, ref SDS 0 2E4H,
310438 -            and indicates linking transforms multiple hierarchies into
310439 -            networks. ref SDS 0 557P
310440 -
310441 -
310442 -
310443 -
3105 -

SUBJECTS
Subject Indexing Purpose Intelligence Support Organize Record Order
What Are We Trying to Accomplish OHS/DKR Collaboration Discussion Gr

5104 -
510501 -  ..
510502 - Purpose Organize Record Find Details Accuraacy Understand Cause Effect
510503 - New Way Working Intelligently Subject Indexing Needs Examples
510504 - Subject Indexing New Way Working Customers Need Examples
510505 - Customers Examples Work Product Functionality Learn New Way Working
510506 -
510507 - Follow up ref SDS B7 PE5T.
510508 -
510509 - Letter to Jack and the guys continues...
510511 -         ..
510512 -    7.  There seems to be agreement among interlocutors the past few
510513 -        days that context management is complex work, ref DIT 1 K69N,
510514 -        reported by Morris on 060905, ref SDS B7 2I7K, and cited by
510515 -        Gary in a letter later the same day. ref SDS B7 JV6Y  There are
510516 -        specific practices and tools using SDS for 8 steps. (see POIMS,
510517 -        ref OF 6 685K) to organize the record of daily work for finding
510518 -        details, assemble chronologies, and construct case studies for
510519 -        intelligence support. (see NWO for scope, ref OF 11 LW5I),
510520 -        ref DIT 1 K69N
510522 -  ..
510523 - SDS tools for subject indexing need examples for people to learn a new
510524 - way of working intelligently, presented in a letter on 060905.
510525 - ref SDS B7 FE5T  People also need support from a librarian type of
510526 - role, also, discussed on 060905. ref SDS B7 HA8I  Morris described
510527 - this work to catelog the record for working efficiently. ref SDS B7
510528 - ML8N  On 010425 described self-evident benefits putting everything in
510529 - the right place at the right time. ref SDS 65 EP7F
510530 -
510531 -
510532 -
510533 -
510534 -
5106 -

SUBJECTS
Federating Model Vison Expand SDS Support Knowledge Management Subje

6803 -
680401 -  ..
680402 - Federating Complex Entailments Needs Explanation Examples
680403 - Experience Doing Knowledge Management Yieds Ideas on Federating SDS
680404 -
680405 - Letter to Jack and the guys concludes...
680406 -
680407 -    8.  This baseline of performance can be enhanced using Jack's
680408 -        vision of a "federating" model, also, presented on 060905.
680409 -        ref SDS B7 PE5T  The challenge, as always, is to perform work,
680410 -        and test results with experience, ref DIT 1 YF4K, keeping in
680411 -        mind that subject indexing to organize the record is one part
680412 -        of intelligence support, listed in POIMS, ref OF 3 0367, and
680413 -        all of the parts have to work together harmoniously. (see again
680414 -        POIMS, ref OF 4 8559), ref DIT 1 YF4K
680416 -  ..
680417 - On 001126 the OHS/DKR team needed experience doing knowledge
680418 - management in order learn what to create for making this work faster
680419 - and easier. ref SDS 58 M171
680420 -
680421 -            [...below Cliff Joslyn presents various data structures,
680422 -            ref SDS 0 2E4H, with comments on viability for application
680423 -            and efficiency, ref SDS 0 556T, evaluation on applying to
680424 -            SDS requires testing. ref SDS 0 3Y4M
680425 -
680426 -
680427 -
680428 -
680429 -
6805 -

SUBJECTS
Subject Index Cross-reference Records Management Filing Context Mana

7803 -
780401 -  ..
780402 - Subject Indexing Cross-references Outlining Hierarchies
780403 -
780404 - Follow up ref SDS B7 DJ9J. ref SDS B5 2I7K.
780405 -
780406 - Received ref DRT 1 0001 from Gary adding to his support received on
780407 - 060905, ref SDS B7 DJ9J, and responding to the letter this morning,
780408 - ref DIT 1 0001, per above, ref SDS 0 QM9I, and to correspondence from
780409 - Jack on 060905. ref SDS B7 QM9I
780411 -  ..
780412 - Gary says...
780413 -
780414 -    1.  Hierarchy is clearly inadequate to index complex bodies of
780415 -        knowledge.  We use hyperlinks to connect the basic data to
780416 -        other basic data, thus creating a network or web.  The same
780417 -        sort of cross-linking needs to be available in our indexes.  I
780418 -        believe that Topic Maps supports this, and it turns out that
780419 -        SDS supports the idea of cross-reference topics in a Subject
780420 -        Index page, which I hadn't known. ref DRT 1 0001
780422 -  ..
780423 - Cross referencing for relational support of hierarchial structures in
780424 - the subject index, was reviewed on 960322, ref SDS 24 KR6M, and was
780425 - demonstrated this morining, during a telephone call with Gary.
780426 -
780427 -            [...below, Cliff Joslyn cites methods for hierarchy data
780428 -            objects modeled as DAGs, ref SDS 0 2E4H, and indicates
780429 -            linking transforms multiple hierarchies into networks.
780430 -            ref SDS 0 557P
780432 -  ..
780433 - Design criteria for subject indexing was discussed in a letter on
780434 - 060905. ref SDS B7 DJ9J
780436 -  ..
780437 - Constructing trails of associations cross-referencing a Subject Index
780438 - makes filing, records management, and context management effective
780439 - with a lot of hard work, was reviewed on 060905, ref SDS B7 7R4L,
780440 - essential to maintain order that yields the power of knowledge to
780441 - control the future, further discussed on 060805. ref SDS B7 X73H
780442 -
780443 -            [On 070125 Gary described advantages of SDS subject
780444 -            indexing. ref SDS C0 YP7N
780446 -  ..
780447 - Gary lists hierarchies in the scope of critical capabilities for
780448 - effective subject indexing. ref SDS 0 IB6Y
780450 -  ..
780451 - Gary's letter continues...
780452 -
780453 -    2.  I have found over the years that I need an outline almost
780454 -        everywhere I need a list, and it may well be that I really need
780455 -        a web everywhere I need an outline. ref DRT 1 T461
780457 -  ..
780458 - Gary presented advantages of outlining in a letter on 041111.
780459 - ref SDS A2 3K5N  The next day on 041112 he explained technology
780460 - features that make outlining productive based on experience using
780461 - Maxthink. ref SDS A3 OR6X  This year on 060412 Gary submitted an
780462 - article on Wizfolders that listed benefits of outlining software
780463 - compared to wordprocessing. ref SDS A9 IG8X
780465 -             ..
780466 -            [...below, Cliff Joslyn explains outlining data structures
780467 -            as "trees," and in relation to hierarchies and networks.
780468 -            ref SDS 0 556T
780470 -  ..
780471 - Outlining fits the model of flexible structure explained in the record
780472 - on 890523. ref SDS 1 T18R  Line numbers support outlining illustrated
780473 - by traditional legal briefing, and in SDS records, also, reviewed on
780474 - 890523. ref SDS 1 T1VQ
780476 -  ..
780477 - Shallow outlining was discussed again on 020217. ref SDS 74 3105
780479 -  ..
780480 - Gary says...
780481 -
780482 -    3.  I really believe in multiple hierarchies, as in multiple
780483 -        inheritance in programming languages, and the categories in
780484 -        Iris.  This is similar to facets in thesaurus work.  This
780485 -        results in multiple "primary" indexes as is done with the 3
780486 -        card catalogs in a library.  Of course, a single-rooted
780487 -        hierarchy can simulate a multi-rooted hierarchy just 1 level
780488 -        down, so for links it doesn't matter how the tree (or forest)
780489 -        is built.  In inheritance, the problem is a bit more complex.
780490 -        ref DRT 1 T465
780491 -
780492 -            [...below, Cliff Joslyn cites hierarchy data objects
780493 -            modeled as DAGs, ref SDS 0 2E4H, and indicates linking
780494 -            transforms multiple hierarchies into networks. ref SDS 0
780495 -            557P
780497 -  ..
780498 - Gary's reliance on 3-card catalogs in library work, follows up
780499 - analysis on 060905, ref SDS B7 JV8P, to illustrate success applying
780500 - multiple hierarchies for indexing data structures aligns with
780501 - solutions for the sharing model Morris raised on 060905. ref SDS B7
780502 - LU56
780504 -  ..
780505 - Morris has sometimes described SDS support for "3x5 card" organizing
780506 - systems, for example on 950223. ref SDS 16 9933
780508 -  ..
780509 - Relational and hierarchial design fits the model of "multiple
780510 - hierarchies, illustrated by analogy in the record on 890523.
780511 - ref SDS 1 G14K  Morris described this SDS breakthrough in a letter on
780512 - 910810. ref SDS 4 7793
780514 -  ..
780515 - "Iris" may relate to the letter this morning crediting Jack's work at
780516 - SRI developing systems for managing catagories, per above, ref SDS 0
780517 - K45X, and citing review on 060211. ref SDS A7 6Z5W
780518 -
780519 -
780520 -
780521 -
780522 -
780523 -
7806 -

SUBJECTS
Subject Indexing Natural Structures Organize Complexity Daily Work E

A403 -
A40401 -  ..
A40402 - Email Subject Indexing Natural Catagories Context Management Easier
A40403 - Integrating Information Streams Makes Better Sense of Email
A40404 - Complementarity Organizing Record Makes Sense for Authors and Readers
A40405 - Subject Indexing Natural Structures Organize Complexity Daily Work
A40406 -
A40407 - Follow up ref SDS B7 F66V.
A40409 -  ..
A40410 - Natural Structures Make Sense Complexity Integration Complementarity
A40411 - Automatic Organization Email Helps Make Sense Complex Communications
A40412 -
A40413 - Gary says...
A40414 -
A40415 -    4.  Where the environment provides some natural categories,
A40416 -        automation can indeed be useful.  Indexing a set of emails by
A40417 -        date, recipients, sender, subjects, etc. is quite helpful.  If
A40418 -        this were enhanced by convention in the emails to add
A40419 -        additional context as is done in formal papers, which often...
A40420 -        supply indexing terms, the leverage can be powerful.  Tagging
A40421 -        an email by project(s), and adding indexing terms could help a
A40422 -        lot.  Failing to have that done by the authors of the emails,
A40423 -        the ability for a user to add such information easily when
A40424 -        transferring the email into the system is valuable.  Iris
A40425 -        supports this. ref DRT 1 T473
A40427 -  ..
A40428 - Would be helpful to see a demonstration of Iris support for managing
A40429 - context in email and other information streams.
A40431 -  ..
A40432 - Indexing "Subject" in email, ref SDS 0 M97O, is problematic.  People
A40433 - rarely enter a useful subject, and often a letter has many subjects.
A40434 - Exhanging correspondence changes the subject, yet people almost never
A40435 - make minimal efforts for Knowledge Mangement to align the Subject with
A40436 - evolving content in email, noted by Jack Park in his letter on 010908,
A40437 - ref SDS 67 UV4H, and later cited by Gary describing the "2-worlds"
A40438 - problem on 030316. ref SDS 85 H69J  Below, Gary describes benefits of
A40439 - Knowledge Management to save lives, time, and money by working
A40440 - intelligently do not seem in the moment comensurate with costs.
A40441 - ref SDS 0 IA5Y
A40443 -  ..
A40444 - Email is a new technology that makes communication the biggest risk in
A40445 - enterprise.  It has elements of accuracy endemic to writing, and of
A40446 - speed that increases productivity, which are both illusory for reasons
A40447 - listed in the record reviewing Jack Park's letter on 010908.
A40448 - ref SDS 67 YF5O
A40450 -  ..
A40451 - Email illusion of productivity was reviewed by OHS/DKR team for adding
A40452 - data structures considered by Gary today, per above, ref SDS 0 IA9X,
A40453 - reported on 000420. ref SDS 44 0784
A40455 -  ..
A40456 - Email makes better sense integrating subject indexing with everything
A40457 - else to augment intelligence.  Integrated information streams improve
A40458 - understanding of daily work by "connecting the dots" for everything
A40459 - into contextual relevance.  Morris mentioned this on 980405.
A40460 - ref SDS 33 FN9F  Earlier, research on 900303 makes a similar point.
A40461 - ref SDS 2 6006
A40462 -
A40463 -
A40464 -
A40465 -
A405 -

SUBJECTS
Subject Index Knowledge Management Command Control Meta Data Sensema

AN03 -
AN0401 -  ..
AN0402 - Knowledge Management Hard Work Construct Assign Subjects
AN0403 - Metadata Assign Subjects Facilitate Finding Retrieval Hard Work
AN0404 - Sensemaking Subject Indexing Context Management Organic Structure
AN0405 - Automatic Subject Indexing Document Content Complex Context Evolves
AN0406 -
AN0407 -
AN0408 - Big Projects Degrade Management to Enropy Technology Cannot Keep Up
AN0409 - People Giving Up Using Technology for Communication Too Complex
AN0410 - Records Management Augments Intelligence High Volume Overwhelming
AN0411 -
AN0412 - Gary's letter continues...
AN0413 -
AN0414 -    5.  I am not certain that setting work context (project, action
AN0415 -        category, etc.) and having that transferred automatically to
AN0416 -        work product created is workable.  I have tried this with
AN0417 -        timekeeping systems with very little luck, since the context
AN0418 -        switches are often frequent and for short periods of time.  It
AN0419 -        would be an interesting experiment to see if a set of context
AN0420 -        parameters could be inherited by activity in something like
AN0421 -        Iris.  Open a data item, create an email or a note, and they
AN0422 -        inherit the context of the data item unless overridden.  Sort
AN0423 -        of like replying to an email carries some context along with
AN0424 -        it. ref DRT 1 T483
AN0426 -  ..
AN0427 - The letter on 060905 reviewed progress on automatic processing to
AN0428 - manage the organic structure of context. ref SDS B7 FF3Y
AN0429 -
AN0430 -            [On 070125 Gary described advantages of SDS subject
AN0431 -            indexing. ref SDS C0 YP7N
AN0432 -
AN0433 -            [On 070511 enhanced context management system to construct
AN0434 -            accounts in the Subject Index and assign to Control Fields.
AN0435 -            ref SDS C2 0001
AN0437 -  ..
AN0438 - SDS Utopia Everything in Right Place at Right Time
AN0439 - Pain Using Management Technology Ignore Bumbling Losing Time Money
AN0441 -  ..
AN0442 - Gary's letter continues...
AN0443 -
AN0444 -    6.  Of course, the indexing and retrieval problem is nothing new.
AN0445 -        One of the intents of XML and the various metadata initiatives
AN0446 -        is to allow authors to add indexing information to data to
AN0447 -        facilitate retrieval.  For a formal document that takes months
AN0448 -        to create, the additional time to add good metadata isn't
AN0449 -        burdensome.  When we ask anyone to do this for short emails,
AN0450 -        notes, reminders, etc., the effort to add the metadata begins
AN0451 -        to meet or exceed that required to do the work, and the
AN0452 -        resistance to doing it goes up, particularly if there is no
AN0453 -        apparent value to it right now.  Being able to find it,
AN0454 -        retrieve it, and correlate it later seldom outweighs the pain
AN0455 -        of having to do it in the moment. ref DRT 1 T493
AN0457 -  ..
AN0458 - Gary's comment that people avoid "pain" of cognitive overhead doing
AN0459 - hard work organizing the record, cited in research at SRI on 000307,
AN0460 - ref SDS 41 767G, and accept greater pain later losing lives, time, and
AN0461 - money, when guess and gossip, hunch, and hope cause continual bumbling
AN0462 - reflects the report by Ray Ozzi, developer of Lotus Notes.  On 020822
AN0463 - Ozzi noted that people feel pain using management technology because
AN0464 - nobody can find critical details in time to be effective. ref SDS 82
AN0465 - 466L
AN0467 -  ..
AN0468 - Gary's comment that records management is "nothing new" aligns with
AN0469 - background on SDS features and benefits.
AN0471 -  ..
AN0472 - Research on 970707 shows management degrades to entropy, i.e., guess
AN0473 - and gossip, hunch and hope in endless meetings, calls, and documents,
AN0474 - because order cannot be maintained in the record for finding critical
AN0475 - details in time to be effective.  People pay a price relying on
AN0476 - remembering only the gist of the story, ref SDS 30 0108, citing review
AN0477 - on 950204, ref SDS 14 0550, and research on 900303. ref SDS 2 6689
AN0478 - Kissinger noted accuracy degrades into a comedy of errors, reported on
AN0479 - 940609. ref SDS 8 4238  Continual bumbling eventually reaches
AN0480 - critical mass and explodes into loss of life, time, and money, e.g.,
AN0481 - breakdown of national security on 010911, ref SDS 68 UP5K, collapse of
AN0482 - Enron into bankruptcy, 020204, ref SDS 73 4N6F, Columbia space shuttle
AN0483 - explodes returning from space, because NASA management interpretted
AN0484 - requirements for good management as suggestions which could be ignored
AN0485 - with good communication skills, reported to Congress on 030826.
AN0486 - ref SDS 90 8K4G
AN0488 -     ..
AN0489 -    Technology Augment Intelligence Records Management
AN0490 -    History Records Management Bush Proposed Trails Associations
AN0491 -
AN0493 -  ..
AN0494 - Vannevar Bush led the nation's research for developing weapons systems
AN0495 - to win World War II.  Problems working with mountains of information
AN0496 - from daily meetings, calls, and documents led to a proposal in 1945 on
AN0497 - using technology for constructing trails of associations to maintain
AN0498 - order in the record of daily work so that finding critical details,
AN0499 - correlations, implications, and nuance would be fast and easy,
AN0500 - reviewed on 960304. ref SDS 22 L47F
AN0502 -     ..
AN0503 -    Roman Empire History Records Management
AN0504 -
AN0505 -
AN0506 - Records management dates back to at least the Roman Empire, noted in
AN0507 - research...
AN0508 -
AN0509 -               [On 070126 case study Knowledge Management objectives
AN0510 -               using computers for records management have all failed.
AN0511 -               ref SDS C1 QH8L
AN0513 -             ..
AN0514 -            ARMA Records Management Quarterly,
AN0515 -            Oct 1994  by Stephens, David O
AN0517 -               ..
AN0518 -              http://findarticles.com/p/articles/mi_qa3691/is_199410/ai_n8717832
AN0519 -
AN0520 -            ...the origins of records management... date back at least
AN0521 -            2,000 years, to the days of...the Roman Empire.
AN0523 -             ..
AN0524 -            Romans ...established a sophisticated records management
AN0525 -            infrastructure.  Administrative organizations and systems
AN0526 -            were built upon the legal foundation of these record
AN0527 -            keeping systems.
AN0529 -             ..
AN0530 -            Roman documents created on wooden tablets were joined
AN0531 -            together when they became inactive; a group of related
AN0532 -            documents formed what was called a codex--a collection of
AN0533 -            records similar to our modern-day file folders.  Within
AN0534 -            each codex, individual tablets were numbered in consecutive
AN0535 -            order and indexed for subsequent retrieval.  Large
AN0536 -            quantities of these tablets were stored in a building
AN0537 -            called a tablinum (literally a "house of tablets"), a
AN0538 -            facility similar to our present-day records centers.
AN0540 -             ..
AN0541 -            In the year 79 B.C., the Romans built the Tabularium, a
AN0542 -            facility adjacent to the Forum, which served as a central
AN0543 -            records repository.  The Tabularium was administered by a
AN0544 -            fourteen quasestores, and during the days of the republic
AN0545 -            free access to its holdings was granted to every Roman
AN0546 -            citizen.
AN0548 -             ..
AN0549 -            The Roman empire from 27 BC until 476 AD maintained records
AN0550 -            in repositories called tabularium Caesaris. "Decentralized"
AN0551 -            centers were established in each province, which contained
AN0552 -            the records of both the central and the provincial
AN0553 -            governments, and there were municipal record centers as
AN0554 -            well.
AN0556 -             ..
AN0557 -            After the Roman Empire, Italian city states of the Middle
AN0558 -            Ages made substantial progress in records management
AN0559 -            systems.  The chancery offices were subdivided into various
AN0560 -            bureaus, each maintaining records and files of the local
AN0561 -            government.  Typically, records were filed either by
AN0562 -            subject, by format, or by type of transaction, and then
AN0563 -            chronologically within these groupings.  These bureaus
AN0564 -            utilized armoires or cupboards as filing cabinets, each
AN0565 -            with a label to indicate the contents.
AN0567 -  ..
AN0568 - In the early 1960s Doug Engelbart proposed a project to implement the
AN0569 - Bush vision presented in 1945, reviewed on 991222, ref SDS 38 XS6N,
AN0570 - and suggested that using technology for records of management augments
AN0571 - intelligence. ref SDS 38 7H5H  This application aligns with SDS design
AN0572 - and practices for intelligence support using Communication Metrics,
AN0573 - discussed with Doug on 991222. ref SDS 38 TY6G
AN0575 -  ..
AN0576 - MIT records management practices present conventional approaches to
AN0577 - subject indexing for context management, also, reviewed on 960322.
AN0578 - ref SDS 24 PX54
AN0580 -  ..
AN0581 - On 000623 Jack Park proposed an engine to automatically make sense of
AN0582 - information by assigning categories. ref SDS 49 2915  Jack discussed
AN0583 - this idea again on 001025. ref SDS 57 LR6N
AN0584 -
AN0585 -            [On 070511 developed SDS context management that constructs
AN0586 -            multiple accounts in the Subject Index and assigns to
AN0587 -            Control Fields. ref SDS C2 0001
AN0589 -  ..
AN0590 - Today, Gary sets out reasons for resisting subject indexing to avoid
AN0591 - the burden of cognitive overhead, ref SDS 0 IA5Y, explained in NWO,
AN0592 - ref OF 13 LH6K, and which are balanced by benefits derived from the
AN0593 - power of knowledge.
AN0595 -  ..
AN0596 - Better productivity requires command and control of the record for
AN0597 - quickly finding and retrieving in chronological order critical details
AN0598 - in time to be effective for saving lives, time, and money.  Order is
AN0599 - the underlying mental mechanism that makes sense of complexity by
AN0600 - "connecting the dots" of cause and effect, commonly called "analysis,"
AN0601 - "study," "research," and sometimes "intelligence" that yields the
AN0602 - power of knowledge to control the future, explained in POIMS.
AN0603 - ref OF 3 1X6G  SDS leverages intelligence that manages the organic
AN0604 - structure of context to retrieve information in chronological order by
AN0605 - narrowing the gap between subconscious and conscious processing, also,
AN0606 - explained in POIMS. ref OF 4 6J4M  Gary points out today that people
AN0607 - get discouraged because consciously adding organization to make
AN0608 - meta data effective takes a lot of hard work to perform context
AN0609 - management, ref SDS 0 IA5Y, discussed above, ref SDS 0 6U9K, and
AN0610 - presented in research at SRI on 000307. ref SDS 41 767G
AN0612 -  ..
AN0613 - Experience grows awareness of benefits using tools for organizing to
AN0614 - improve work, discussed on 060905. ref SDS B7 CZ3H  Investing time to
AN0615 - understand "meaning" makes sense of complexity by constructing orderly
AN0616 - trails associating cause and effect based on contextual relevance
AN0617 - managed by the Subject Index.  Accurate understanding of causation
AN0618 - helps write accurate, comprehensive, and concise correspondence,
AN0619 - specifications, and reports, which support planning for taking
AN0620 - effective action, explained in POIMS. ref OF 5 0859
AN0622 -  ..
AN0623 - Jack's seminar on 060523 presented ideas on merging ontologies being
AN0624 - developed and promoted at SRI. ref SDS B0 UK6Q  Another seminar could
AN0625 - demonstrate how Open Iris uses categories and other features that add
AN0626 - context to email, mentioned by Gary today, as part of merging
AN0627 - ontologies.
AN0629 -  ..
AN0630 - The biggest challenge is maintaing order that enables the power of
AN0631 - knowledge to predict the future based on the locality principle,
AN0632 - reviewed on 040312. ref SDS 94 YH4G  SDS supports 8 steps for adding
AN0633 - order that yield Communication Metrics, listed in POIMS. ref OF 6 685K
AN0635 -  ..
AN0636 - Gary explained this challenge in a call on 030316. ref SDS 85 7J7I
AN0638 -  ..
AN0639 - Gary described challenges of organizing documentation with an organic
AN0640 - subject index, in another call on 030408. ref SDS 87 WE7N
AN0642 -  ..
AN0643 - On 030512 Gary discussed scope of 2-worlds, that has many gaps to
AN0644 - transcend from information to a culture of knowledge. ref SDS 88 RX5O
AN0646 -  ..
AN0647 - Gary's idea to hire a "refractor" to catelog the record shifts pain to
AN0648 - gain by paying people to organize everything.  For example, a
AN0649 - librarian never worries about extra time to catelog, because the more
AN0650 - there is to catelog, the more they get paid, reviewed on 060905.
AN0651 - ref SDS B7 KC8H
AN0653 -  ..
AN0654 - Jack Park proposed in a letter on 040622 using complexity theory for
AN0655 - sensemaking. ref SDS 98 FM5W  Review showed that people make sense of
AN0656 - complexity by constructing trails of correlations and implications,
AN0657 - and discovering nuance that is otherwise hidden by limited span of
AN0658 - attention, especially when time is short. ref SDS 98 T66N
AN0660 -  ..
AN0661 - Gary's explains writers perform Knowledge Management to make sense of
AN0662 - writing with chapter descriptions that summarize meaning (sometimes),
AN0663 - creating a table of contents for navigation, and by indexing to locate
AN0664 - content.  Research at SRI on 000307 found that Knowledge Management a
AN0665 - lot of hard work which discourages performance, ref SDS 41 767G, as
AN0666 - Gary notes today. ref SDS 0 IA5Y  Readers have complementary work to
AN0667 - associate content with objectives, requirements, and commitments under
AN0668 - MBO criteria, reported on 960322. ref SDS 24 TS04, and citing earlier
AN0669 - analysis on 890523. ref SDS 1 XT6O  Binary structure explained in NWO,
AN0670 - ref OF 10 VL4G, requires mutual efforts by authors and readers to
AN0671 - construct meaning by associating subjects with complementary
AN0672 - objectives.  Complementarity is reviewed in the record on 040312.
AN0673 - ref SDS 96 486I
AN0674 -
AN0675 -     [On 061102 example 2-worlds problem using 8-steps of Communication
AN0676 -     Metrics to make sense of complex medical report on coagulapathy
AN0677 -     report. ref SDS B9 6R4O
AN0679 -  ..
AN0680 - Natural structures that reduce the burden of cognitive overhead for
AN0681 - subject indexing were reviewed for Gary's letter on 060905.
AN0682 - ref SDS B7 F66V  Examples of natural hierarchies for subject indexing
AN0683 - are listed in the record on 960322. ref SDS 24 FG4M  The record on
AN0684 - 890523 illustrates natural structure with WBS for a "big picture."
AN0685 - ref SDS 1 428F
AN0686 -
AN0687 -
AN0688 -
AN07 -

SUBJECTS
Interoperability Common Storage Enable Knowledge Management Compleme

B703 -
B70401 -  ..
B70402 - Interoperability Common Storage Enable Knowledge Management
B70403 - Meta Data Index Finding Information Requires Common Storage
B70404 - Common Storage Interoperability Complement Subject Index Meta Data
B70405 -
B70406 -
B70407 - Gary's letter continues...
B70408 -
B70409 -    7.  Both SDS and Iris recognize that integration requires some form
B70410 -        of common data store, and interaction with the applications
B70411 -        that create and manage the data items.  Addressing emails and
B70412 -        documents in their native applications if far from simple.  It
B70413 -        is challenging enough when you transfer the information into a
B70414 -        form that you can control. ref DRT 1 T444
B70416 -  ..
B70417 - SDS support for Knowledge Space applies "common storage," required for
B70418 - Intelligence support, shown in NWO. ref OF 11 PX6J
B70420 -  ..
B70421 - Gary commented on 010408 that records management is a lot of hard work
B70422 - using email for collaboration. ref SDS 63 9H8H  On 030407 Gary listed
B70423 - about 17 software programs needed for managing daily work without SDS.
B70424 - ref SDS 86 234G  This aligns with research on 000307 at SRI by the ad
B70425 - hoc OHS/DKR project team finding that knowledge management is a lot of
B70426 - hard work, ref SDS 41 767G, also, cited above for defining,
B70427 - constructing, and assigning subjects. ref SDS 0 K55F
B70429 -  ..
B70430 - Earlier on 030316 Gary cited "2-worlds" problem.  Variability in
B70431 - document construction tools makes investing time to create a common
B70432 - format for common storage and "connecting the dots" to work
B70433 - intelligently by adding alignment to verify accuracy and discover
B70434 - correlations, implications and nuance takes a lot of hard work that
B70435 - people avoid. ref SDS 85 7J79  An example is PDF formatting reviewed
B70436 - on 061102. ref SDS B8 6592
B70438 -  ..
B70439 - Gary's construction associating "control" with "information" aligns
B70440 - with "command and control"  presented in POIMS for technology to
B70441 - strengthen management of daily work. ref OF 7 1113  Command and
B70442 - control of the record is the missing ingredient that prevents people
B70443 - from adding links to email, reported on 020820, ref SDS 80 UV6G, and
B70444 - further leads people to declare that accuracy is not critical in
B70445 - email. ref SDS 81 V66I
B70447 -  ..
B70448 - This record illustrates managing content in a letter from Gary, and
B70449 - showing correlations with prior letters from Gary and from others.
B70451 -  ..
B70452 - Gary's letter continues...
B70453 -
B70454 -    8.  I am looking at the indexing, metadata issues again to manage
B70455 -        the increasing volume of file on my personal system.  As the
B70456 -        drive capacity goes up, I simply save more stuff.  With only a
B70457 -        single index, the directory path, I can do only so much.  In
B70458 -        theory, I could build alternate paths using links in the
B70459 -        filesystem, but that isn't as easy as I would like on Windows.
B70460 -        I have given any real thought to trying that. ref DRT 1 T445
B70462 -         ..
B70463 -    9.  Using a good checksum or digest gives me a way of locating
B70464 -        duplicate files, which is nice, but if I add metadata to the
B70465 -        files, I destroy that ability unless I arrange to have only a
B70466 -        single copy of the file and update the single copy (I still
B70467 -        can't recognize a duplicate of the original if I should
B70468 -        download another copy, which happens often).  The fact that
B70469 -        metadata structures differ by file type, and the file formats
B70470 -        (mostly Microsoft) are proprietary just adds to the problem.  I
B70471 -        *could* tackle the problem with some external data store for
B70472 -        the index, but with files numbering in the high 10s of
B70473 -        thousands, I am hesitant to start down that road. ref DRT 1
B70474 -        T457
B70475 -
B70476 -
B70477 -
B705 -

SUBJECTS
Explicit Links Granular Addressability Case Study Enable Precision A

BD03 -
BD0401 -  ..
BD0402 - Case Study Explicit Links Granular Addressability Knowledge Management
BD0403 - Granular Addressability Explicit Links Enable Precision Access
BD0404 - Connections Complement Chronology Context with Granular Addressability
BD0405 -
BD0406 -
BD0407 -   10.  Adding granular addressability so that a data item can be
BD0408 -        composed of multiple segments which can be individually indexed
BD0409 -        simply multiplies the problem beyond measure. ref DRT 1 T469
BD0411 -  ..
BD0412 - Explicit links enable granular addressability explained in the scope
BD0413 - for intelligence support in NWO, ref OF 11 6A4K, and provide precision
BD0414 - access to relevant background and authorities.
BD0416 -  ..
BD0417 - Gary has commented previously...
BD0418 -
BD0419 -        1.  Explicit links for granular
BD0420 -            addressability improve
BD0421 -            improve communication............. 020325, ref SDS 76 0001
BD0422 -
BD0423 -        2.  SDS design for explicit links to
BD0424 -            apply granular addressability makes
BD0425 -            precision access fast and easy;
BD0426 -            attitudes changes - people suddenly
BD0427 -            prefer good management to save
BD0428 -            lives time and
BD0429 -            money............................. 020618, ref SDS 78 BK5L
BD0431 -             ..
BD0432 -        3.  Granular addressability valuable
BD0433 -            to users for precision access
BD0434 -            in documents...................... 020618, ref SDS 78 M23V
BD0436 -             ..
BD0437 -        4.  Links annoy people until they gain
BD0438 -            experience applying the power of
BD0439 -            knowledge derived from connecting
BD0440 -            cause and effect.................. 020730, ref SDS 79 5K5O
BD0442 -             ..
BD0443 -        5.  Case study power of links to make
BD0444 -            useful connections, and fear people
BD0445 -            suffer until there is experience
BD0446 -            using the power of
BD0447 -            knowledge......................... 030524, ref SDS 89 GE4N
BD0449 -             ..
BD0450 -        6.  Explicit links and granular addressability
BD0451 -            presented as "hyperlinks" in a
BD0452 -            letter on SDS requested by
BD0453 -            Jack Park......................... 040301, ref SDS 93 ML74
BD0455 -             ..
BD0456 -        7.  2-worlds problem connecting to wide
BD0457 -            range data formats applied in
BD0458 -            SDS adding explicit links for
BD0459 -            granular addressability to enable
BD0460 -            command and control of the
BD0461 -            record............................ 040301, ref SDS 93 ML8V
BD0463 -             ..
BD0464 -        8.  Weblogs and blogging popular efforts
BD0465 -            using the Internet for communication
BD0466 -            fails to support power of knowledge
BD0467 -            from explicit links for grannular
BD0468 -            addressability.................... 040802, ref SDS 99 ME5N
BD0470 -             ..
BD0471 -        9.  SDS explicit links and granular
BD0472 -            addressability enable precision
BD0473 -            access presented for ASB
BD0474 -            brief description................. 041025, ref SDS A0 JN9K
BD0476 -             ..
BD0477 -       10.  Explicit links, granular addressability
BD0478 -            and precision access explained
BD0479 -            in traditional terms of
BD0480 -            documentation for better
BD0481 -            accuracy.......................... 041108, ref SDS A1 B46G
BD0483 -             ..
BD0484 -       11.  Chronology, context, and connection
BD0485 -            summarize SDS support for working
BD0486 -            intelligently using explicit links
BD0487 -            for granular addressability that
BD0488 -            gives SDS the power of
BD0489 -            knowledge......................... 051021, ref SDS A5 VI5U
BD0491 -             ..
BD0492 -            Case study Gary explaining benefits
BD0493 -            explicit links for granular
BD0494 -            addressability.................... 060907, ref SDS 0 IB5V
BD0496 -  ..
BD0497 - Experience adding granular addressability shows that organizing
BD0498 - complexity reduces problems that result from rising entropy when
BD0499 - energy is not added to connect the order of cause and effect, reviewed
BD0500 - on 040312, ref SDS 94 YH4G, and discussed above citing Vannevar Bush
BD0501 - proposing in 1945 technology to construct trails of associations that
BD0502 - help manage complexity in the organic structure of context.
BD0503 - ref SDS 0 665K
BD0505 -  ..
BD0506 - Complexity that is organized increases productivity. ref SDS 95 5P9I
BD0508 -  ..
BD0509 - Complexity without granular structuring of experience based on context
BD0510 - devolves into a liability that paralyzes productivity. ref SDS 95 W46M
BD0511 -
BD0512 -
BD0513 -
BD0514 -
BD0515 -
BD06 -

SUBJECTS
Subject Index Requirements Summarized Topic Maps Categories Records

BJ03 -
BJ0401 -  ..
BJ0402 - Subject Indexing Requirements Summarized
BJ0403 -
BJ0404 - Gary seems to describe in this part of his letter today the scope of
BJ0405 - SDS subject indexing reported in the record on 950626, ref SDS 19
BJ0406 - 2882, citing prior work on 890523. ref SDS 1 T15F  Analysis on 960322
BJ0407 - also has helpful perspective on granular division of accounts,
BJ0408 - ref SDS 24 PX7F, and difficulties of transitions. ref SDS 24 B25I
BJ0410 -  ..
BJ0411 - This could be incorporated into Morris' scope for SDS to expand
BJ0412 - explanation on 950221 of the subject index. ref SDS 15 MT9F
BJ0414 -  ..
BJ0415 - Gary's letter continues...
BJ0417 -         ..
BJ0418 -   11.  Some elements are clear from experience:
BJ0419 -
BJ0420 -            1.  The index has to be separate from the data indexed.
BJ0421 -                ref DRT 1 T532
BJ0423 -                 ..
BJ0424 -            2.  There has to be more than 1 primary index (or you could
BJ0425 -                just structure the data store). ref DRT 1 KP4G
BJ0427 -                 ..
BJ0428 -            3.  The index structure has to be at least hierarchic.
BJ0429 -                ref DRT 1 1Q4J
BJ0430 -
BJ0431 -                    [On 070125 Gary described scope and advantages of
BJ0432 -                    SDS subject indexing. ref SDS C0 YP7N
BJ0434 -                 ..
BJ0435 -            4.  Better is to have the index use multiple hierarchies
BJ0436 -                (facets?). ref DRT 1 TQ4L
BJ0438 -                 ..
BJ0439 -            5.  Every data item can have multiple categories in various
BJ0440 -                places in the index. ref DRT 1 FR4N
BJ0442 -                 ..
BJ0443 -            6.  It is valuable to be able to express the same indexing
BJ0444 -                term in a variety of ways, similar to the Preferred
BJ0445 -                Term concept in a thesaurus. ref DRT 1 SR5G
BJ0447 -                 ..
BJ0448 -            7.  In an advanced system, the index itself needs to
BJ0449 -                support hyperlinks between and among index elements.
BJ0450 -                ref DRT 1 6S5J
BJ0452 -                 ..
BJ0453 -            8.  There is value to being able to replicate a
BJ0454 -                sub-structure in different contexts.  Projects of the
BJ0455 -                same type have similar indexing structures at the top
BJ0456 -                level, but the fact that the project is different
BJ0457 -                provides slightly different meaning to the subjects, so
BJ0458 -                a link isn't what is wanted.  This is where our
BJ0459 -                discussion began - is there value in providing SDS
BJ0460 -                users with Rod's populated Subject Index. ref DRT 1
BJ0461 -                U45M
BJ0463 -                 ..
BJ0464 -            9.  The idea of a unique description of a topic as used in
BJ0465 -                Topic Maps is quite nice. ref DRT 1 OT6J
BJ0467 -  ..
BJ0468 - Uniqueness within a flexible organic structure of context combines
BJ0469 - benefits of relational and hierarchial design, set out in planning on
BJ0470 - 890523. ref SDS 1 T15F  This was explained on 031207. ref SDS 91 G27M
BJ0471 -
BJ0472 -
BJ0473 -
BJ0474 -
BJ0475 -
BJ05 -

SUBJECTS
Subject Indexing Complex Task Organize Complexity Daily Work Conside

BW03 -
BW0401 -  ..
BW0402 - Critical Mass Tools Hard to Find to Begin Organizing the Work
BW0403 - Email Incremental Step to Knowledge Management Leverage Intelligence
BW0404 -
BW0405 - Gary's letter continues...
BW0406 -
BW0407 -   12.  The total package is huge.  As yet, I haven't seen a good
BW0408 -        incremental approach as a "critical mass" of capability is
BW0409 -        necessary in order for the system to begin being part of the
BW0410 -        solution space rather than part of the problem space.  Or,
BW0411 -        rather, I haven't found a good incremental approach that can be
BW0412 -        done with my resources.  But, I keep thinking about it and
BW0413 -        looking at it.  As other build some of the pieces, the reality
BW0414 -        may grow closer.
BW0416 -  ..
BW0417 - The scope of "huge package" for "critical mass" of capability cited in
BW0418 - this part of Gary's letter seems related to Gary's list above,
BW0419 - ref SDS 0 IB6Q, which generally describes the SDS subject index design
BW0420 - for managing of daily work by creating order and structure to
BW0421 - understand complexity, per above. ref SDS 0 Y94I  What steps
BW0422 - additional tip the balance toward organizing the record?  Is this the
BW0423 - librarian discussed on 060905? ref SDS B7 KC8H
BW0425 -  ..
BW0426 - Gary seems to present today context management as hopelessly beyond
BW0427 - reach, repeating the letter on 060905, ref SDS B7 DJ9J, that aligns
BW0428 - with Jack's letter on 000221 citing authority on complexity of
BW0429 - organic subject structure, like Pandora's Box.
BW0430 -
BW0431 -            [On 070125 Gary described advantages of SDS subject
BW0432 -            indexing. ref SDS C0 YP7N
BW0434 -  ..
BW0435 - In some respects, Gary reviews issues today considered by Jack Park
BW0436 - and others in 2000 setting objectives for the OHS/DKR, shown by Doug
BW0437 - Engelbart's letter on 000405 proposing to first add addressability to
BW0438 - email for linking to cite the record. ref SDS 43 2484  At that time,
BW0439 - on 000426 Jack wrote...
BW0441 -         ..
BW0442 -        Thus far, Doug has elucidated a rather vast and huge picture of
BW0443 -        what needs to be done (his vision), but has not yet drilled
BW0444 -        down to the bloody details of what that should look like.
BW0445 -        Right now, we know we want a dynamic knowledge repository.
BW0446 -        What the hell is that? ref SDS 45 2340
BW0448 -  ..
BW0449 - On 000503 Eric Armstrong reported difficulty with the conceptual leap
BW0450 - from information to "knowledge," ref SDS 46 5033, noted earlier by
BW0451 - Morris on 960227. ref SDS 21 0022  Eric therefore proposed making
BW0452 - improving email the primary step, ref SDS 46 F82E, with features
BW0453 - similar to Gary's list today. ref SDS 0 IA4X and, ref SDS 0 IA5Y  On
BW0454 - 000505 Eric applied this idea in a specification for the OHS/DKR to
BW0455 - use email that improves collaboration as an incremental approach to
BW0456 - Knowledge Management. ref SDS 48 4392
BW0458 -  ..
BW0459 - At that time, on 000503 Jack Park disagreed with Eric, and said...
BW0460 -
BW0461 -        I want the ability to roam around in knowledge space, looking
BW0462 -        for interesting ideas, things others didn't think of.  Can't do
BW0463 -        it with an email system.  Sorry. ref SDS 46 6138
BW0465 -  ..
BW0466 - A few months later on 000824, Eric reported on meeting to review
BW0467 - OHS/DKR architecture, and that Doug Engelbart proposed the SDS system
BW0468 - using links as part of Knowledge Space. ref SDS 54 OU60  Eric reported
BW0469 - on 001127 that he encountered difficulties creating his approach to
BW0470 - better management. ref SDS 59 QI4O
BW0472 -  ..
BW0473 - Subsequently, Jack developed Nexist, and now Open Iris reviewed on
BW0474 - 060211. ref SDS A7 0001  On 010131, Eugene Kim developed tools for
BW0475 - explicit links that provide granular addressability in email,
BW0476 - ref SDS 62 MM9M, based on a format using "purple numbers" demonstrated
BW0477 - by Doug's OHS/DKR launch report on 001025. ref SDS 57 00VU  A year
BW0478 - later, on 020224 Eugene reported improvements to addressability in
BW0479 - email. ref SDS 75 XW9N  The following year, Eugene sponsored a web
BW0480 - site for the OHS/DKR engineers to collaborate using his system of
BW0481 - purple numbers for granular addressability in email, as Doug proposed
BW0482 - on 000405. ref SDS 84 5560  Concurrently, blogging and wikis, and so
BW0483 - on have exploded on the Internet.  Neither Eugene's OHS/DKR group, nor
BW0484 - anbody else have ever added a link.  On 020820 Murray Altheim reported
BW0485 - that links are not useful in email, contrary to all the effort
BW0486 - expended make links possible. ref SDS 80 UV6G  Why not?  Did people
BW0487 - discover that adding connections did not improve accuracy, memory, and
BW0488 - leverage intelligence, as expected?  No!
BW0490 -  ..
BW0491 - People discovered that creating a connected record requires more than
BW0492 - granular addressability and linking technologies, explained by Morris
BW0493 - on 000706. ref SDS 50 K693
BW0495 -  ..
BW0496 - Netscape, Microsoft, and others have provided some level of control
BW0497 - for email which people use for collaboration.  SDS provides additional
BW0498 - command and control of content for everything, illustrated by the
BW0499 - example on 890523. ref SDS 1 428F  Command and control of the record
BW0500 - is demonstrated by this and other SDS records.
BW0502 -  ..
BW0503 - There are about 2,000,000 pointers in the SDS subject index, reported
BW0504 - on 060905 discussing Morris' comments on SDS design. ref SDS B7 L56G
BW0505 - Subject indexing is one of seven methods for command and control of
BW0506 - the record in Knowledge Space, as defined in NWO. ref OF 11 PX6J
BW0507 -
BW0508 -        1.  Chronology
BW0509 -        2.  Links
BW0510 -        3.  References
BW0511 -        4.  Document Log
BW0512 -        5.  Contacts
BW0513 -        6.  Functions
BW0514 -        7.  Subjects
BW0516 -  ..
BW0517 - These are complementary capabilities for people to find ideas,
BW0518 - critical details, and trends of cause and effect quickly and easily,
BW0519 - without roaming around, and in time to be effective getting things
BW0520 - done.  Part of doing the work is building a connected record that
BW0521 - strengthens accuracy to avoid mistakes, and grows knowledge for
BW0522 - solving problems by constructing an external rendering of internal
BW0523 - human thought that can then be shaped and refined and cited.
BW0524 -
BW0525 -
BW0526 -
BW06 -

SUBJECTS
Subject Index Data Structures Hierarchies Data Objects Directed Acyc

C603 -
C60401 -  ..
C60402 - Joslyn Explains Data Structures for Knowledge Hierarchies Networks
C60403 - Context Organic Structures Hierarchies Trees DAGS Latices Outlines
C60404 -
C60405 - Received ref DRT 2 0001 commenting on Gary's letter today, and
C60406 - recognizing participation by Jack, and Morris.
C60408 -  ..
C60409 - Cliff responds to the letter this morning citing his presentation at
C60410 - SRI, ref SDS 0 K46T, following Jack's letter on 060905 indicating
C60411 - Cliff presented hypergraphs. ref SDS B7 W955
C60412 -
C60413 -    1.  Very interesting. I would make these observations:
C60414 -
C60415 -        1.  "Hierarchy" does not mean "tree".  Hierarchies are properly
C60416 -            modeled in terms of data objects as DAGs, and
C60417 -            mathematically as posets and/or lattices.  These are rich
C60418 -            in multiple inheritance, and have distinct levels, but
C60419 -            those levels can be of different sizes, and can themselves
C60420 -            overlap. ref DRT 2 0001
C60422 -  ..
C60423 - Apparently DAGs is an acronym for...
C60424 -
C60425 -                     directed acyclic graph
C60426 -
C60427 -    Free patents online
C60428 -
C60429 -    Method and system to compare data objects
C60430 -    United States Patent 20060161560
C60431 -
C60432 -              http://www.freepatentsonline.com/20060161560.html
C60433 -
C60434 -        A directed acyclic graph (DAG) is a representation of a set of
C60435 -        items, each of which is associated with a node of the graph.
C60436 -        All nodes of a directed acyclic graph are connected by edges or
C60437 -        logical connections which are unidirectional (i.e. directed) in
C60438 -        nature.  Further, a route traced along connected edges, in the
C60439 -        direction specified by the edges, never ends on a node from
C60440 -        which the route starts (i.e. acyclicity). [0029]
C60442 -         ..
C60443 -        Directed acyclic graph forest:
C60445 -         ..
C60446 -        A directed acyclic graph forest is a set of one or more
C60447 -        directed acyclic graphs. [0030]
C60449 -         ..
C60450 -        Ontology-based attribute:
C60452 -         ..
C60453 -        Ontology-based attribute is an attribute that has a multi-level
C60454 -        description.  Every element in a level is related to its
C60455 -        ancestors by certain defined relationships.  An ontology based
C60456 -        attribute may be represented as a DAG structure. [0031]
C60458 -         ..
C60459 -        Simple attribute:
C60461 -         ..
C60462 -        Simple attribute is an attribute that can have only one value.
C60463 -        Simple attribute may also be defined as an ontology-based
C60464 -        attribute with only one node. [0032]
C60466 -         ..
C60467 -        Ancestor nodes:
C60469 -         ..
C60470 -        In a DAG representation of data, for any given node, ancestor
C60471 -        nodes include the nodes that fall on a path traversed from the
C60472 -        root node of the [0033] DAG to the given node.  Root node is
C60473 -        always included in the set of ancestor nodes. [0034] Descendent
C60474 -        nodes:  In a DAG representation of data, for a given node,
C60475 -        descendent nodes include the nodes that fall on a path
C60476 -        traversed from the given node to all leaf-nodes connected to
C60477 -        the given node.  All leaf-nodes connected to the given node are
C60478 -        included in the set of descendent nodes. [0035] Parent nodes:
C60479 -        In a DAG representation of data, parent nodes of a given node
C60480 -        are the immediate ancestor nodes of the given node. [0036]
C60481 -        Child nodes:  In a DAG representation of data, child nodes of a
C60482 -        given node are the immediate descendent nodes of the given
C60483 -        node.
C60485 -  ..
C60486 - On 060905 Gary's letter discussed challenges achieving an effective
C60487 - sharing model using complex hierarchy data structures for context
C60488 - management ref SDS B7 DJ9J, concurring with concerns presented by
C60489 - Morris that subject indexing requires study to apply for making
C60490 - finding critical details fast and easy. ref SDS B7 2I7K  Cultural
C60491 - practices use specialists to help people apply complex hierarchies,
C60492 - for example at the library. ref SDS B7 LU56
C60494 -  ..
C60495 - Today, Gary says hierarchies are inadequate for managing complex
C60496 - records ("bodies of knowledge") ref SDS 0 9J3N, and indicates multiple
C60497 - hierarchies offer a better solution, and credits Open Iris for
C60498 - implementation. ref SDS 0 IA8Y
C60499 -
C60501 -  ..
C60502 - Cliff's letter continues...
C60503 -
C60504 -        2.  Trees are highly restricted and impoverished cases of
C60505 -            hierarchies.  Outlines are technically trees.  Note that
C60506 -            ordered lists are technically also hierarchies, although
C60507 -            even MORE restricted and impoverished. ref DRT 2 Q76J
C60509 -  ..
C60510 - Outlines are cited in Gary's letter today on data structures for
C60511 - organizing the record by adding order to manage complexity, per above.
C60512 - ref SDS 0 IA8U
C60514 -  ..
C60515 - Cliff's letter continues...
C60516 -
C60517 -        3.  As soon as you introduce arbitrary hyperlinks which can
C60518 -            cross arbitrary levels, you've moved to a full network
C60519 -            model.  That's it, there's nothing in between: you're
C60520 -            either hierarchical or you're a full network (directed
C60521 -            graph).  There's not much value in introducing intermediate
C60522 -            concepts or novel terms like "heterarchy" or "web".
C60523 -            ref DRT 2 N46O
C60525 -  ..
C60526 - "Network Centric Organization" NCO has become a prominent term
C60527 - replacing Knowledge Management, "Business Intelligence," etc., for
C60528 - marketing products and services, reviewed on 980228. ref SDS 31 2716
C60529 - An example occurrs from discussions with Boeing on 060317. ref SDS A8
C60530 - IT5V
C60532 -  ..
C60533 - Cliff seems to address Jack's question on 060905 about a presentation
C60534 - at SRI on hypergraphs and heterarchies, ref SDS B7 W955, and which was
C60535 - presented by copy to Cliff today, per above. ref SDS 0 K47P
C60537 -  ..
C60538 - Cliff's letter continues...
C60539 -
C60540 -        4.  Unions of hierarchies (facets), one for each link and/or
C60541 -            node type present, are clearly the proper way to organize
C60542 -            complex knowledge systems.  I believe that the concepts of
C60543 -            "multiple hierarchies" and "multiple inheritance" are
C60544 -            actually distinct, as single hierarchies can have multiple
C60545 -            inheritance. ref DRT 2 N47U
C60547 -             ..
C60548 -        5.  But in a facted system, the crucial questions which arise
C60549 -            are as to whether those unions of hierarchies (say, taken
C60550 -            pairwise) are themselves acyclic (and thus hierarchies), or
C60551 -            whether you can traverse cycles by going "up" one link type
C60552 -            and "down" another.  Different, interesting things happen
C60553 -            one way or the other. ref DRT 2 N48Q
C60555 -  ..
C60556 - Need examples applying data structure alternatives presented in
C60557 - Cliff's letter today for managing typcial SDS records, like the record
C60558 - today, ref SDS 0 0001, the record on 060317, ref SDS A8 0001, or on
C60559 - 060825. ref SDS B4 0001
C60561 -  ..
C60562 - Need for testing reflects the letter to the team early this morning.
C60563 - ref SDS 0 I26M
C60565 -  ..
C60566 - How quickly can such systems be used to develop effective structure,
C60567 - and what kind of benefits are derived for the investment, compared to
C60568 - current methods?
C60570 -  ..
C60571 - Cliff's letter continues...
C60572 -
C60573 -    2.  Happy to expand more on anything of interest here. ref DRT 2
C60574 -        N48X
C60576 -  ..
C60577 - On 000723 Cliff presented interesting explanations of SDS to the group
C60578 - meeting at SRI. ref SDS 52 3960
C60579 -
C60580 -
C60581 -
C60582 -
C60583 -
C60584 -
C60585 -
C60586 -
C60587 -
C60588 -
C606 -