Builder

Bob at 40

Corpus and Site Construction

Plans, Goals, Objectives, Aims, Completed Work Next Review planned in October, 2017
Content Management Systems Usage
This web site is my current attempt to construct a Case Analysis Support Environment. See CASE and The Notion of a Datacase for further discussion. The content will include the information I have collected in my case studies, including the video archives, ascriptions of knowledge states and interpretations of learning experiences, and local theories made about those learning experiences.
It is my further intention to use the facilities of public comment through web logs to foster collaboration with colleagues interested in critical or constructive use of these materials. The objective is to enhance the credibility of case study as a method.

Work In-Process: updating June, 2017 Approaches and Decisions
Revivifying the Agenda:
Review the state of the site:
– decide what is still adequately sound
– decide what needs repair immediately.
– decide what modifications can be done later.
Plan further development:
– what resources need be added or filled out.
– what analyses to pursue, when, and how.
– what developments could be usefully modeled.
Evaluation Control and Planning Issues:
Develop evaluation checklists by indices
– begin with menu specified directories.
– use “immediate repair” criteria to generate task lists.
– use “later modifications” as component of future plans.
Push evaluation into sub-indices as needed.
Earlier Work In-Process: in limbo since research interruption Approaches and Decisions
Progress has begun in modeling by returning to classical AI roots, but the results are modest, limited now by personal and circumstantial constraints. The intention is to employ similar simplification in modeling the contexts and behavior of LC3, the infant Peggy study. COUNT: One Selfridge Model is not yet public, pending revisions in September.
General plan and schedule: method for analyzing video clip threads
Method:DNS echo-transcription to generate episode-texts from videos;
install texts in video clip notes. Link-episodes for analysis
√Method: use digital recorder for read text/commentary.
√Next: use DNS for voice to text processing
√Format & install in Word Press editor.
See first example: LC2bT1 READING SKILLS
√develop method to tag panes (clips) within panels (pages)
~continue video processing, as background activity
-plan on beginning analyses/interpretation early in 2014.
decision:use DNS speech-to-text as input
clipNote formats installed for 11-12th quarter videos: e.g. panel P146
~eg.LC3cA63 Ring-Tower Mastery Thread.
√hardware purchased,software installed:12/2013
~training & testing completed, Jan 2014
Test by reading two pages from CECD.
Formatting embedded in voice record.
HTML added in Word Press installation.
√published in June, 2013;

transient interest
Video Processing and Installation Plans
Files initially uploaded are MP4, 640×480 pixels, 125khz audio; these are good for analysis, but take a long time to load. The original design will be extended to incorporate smaller files (MP4, 320×240 16khz audio) for faster loading. Video panels will load the smaller files; the clipNotes, linked from the panes of the panels, will load the higher resolution files on their pages when a site-visitor activates the link. See Panel P146 and its 7 clipNotes.
The production of smaller video clip files will proceed in parallel with clip analysis and development of clipNotes in the fall of 2013.
Using different browsers with different instances of NLCSA.net is a simple way to see multiple views of one pane or multiple panes and panels.
“5 Pioneers on Learning, AI, & Ed”.
these videos linked from “AItopics.org”.
Development of clipNotes begins now. P146 contains as a structure example a first-pass set of comments for three panes (E, F, and G.).
For now, the best indication of contents is the Video Indices, which contain thematic pointers. See LC3cV01 IPS Video Index and LC2c-Video TIS Video Index.
The archives includes roughly 70 videos of “The Intimate Study” (collected in research at the Logo Project of the MIT Artificial Intelligence Laboratory) and approximately 300 videos of the “Infant Peggy Study” (a six year long weekly chronicle of her development and learning; the first three years were quite regular; the latter three were less so).

The videos have been digitized. The objective is to produce web-accessible clips of the material. (Some reprocessing to improve quality will be necessary.) The videos are presented as “Panels,” edited into episode-clips, “Panes.”
√Backup MySQL db; √install WP 3.6: Sept., 2013.
√plan summer 2013 projects and preparations for fall 2013 
√open video selected materials to a few interested colleagues in Feb. 2013.
√open site to search engine indexing; requires extensive SEO preparation.
√process 3 years of video materials and bring online in Q1, 2013.
√Install Page offprint capability
√Write note for users.
√schedule early June, 2013.

√Chose “printer-friendly” WP Plugin
Some work-arounds required.
√develop Posting Linked Data technique (replacement for inadequate SCPP) √copy Prototype with integrated iframe
√encourage colleagues to explore NLCSA
√Tags for location in Time:
Apply Q# and Y# to all three subjects. Not distinguishing tags for individuals
permits union of data; separation is possible by LC# intersection.
For LC1 and LC2 both, time related to The Intimate Study is important.
Introduce tags for time Before, During and Following the Intimate Study.
weekly tags are not relevant, nor available for LC1 and LC2.
Tag labels ISB, ISD, and ISF chosen to maintain tag sort order in the sidebar list.
√Extend list of tags for subject categories.
Introduce tags AS (Academic Subjects) and CS (Common Sense).
There may be some overlap of these two tag sets.
Introduce tags LT (Logo themes) and FT (Family themes).
Prior subject tags unchanged.
Lng, Soc, SM, Cog, Nbr, Rdg, Cmp
√Tags for relevance categories remain the same for LC1 and LC2. Imp, Evl, Int, Evd: see below.
√Tags for idea categories remain the same for LC1 and LC2. Ntn, Idea, MT, Jdg: see below.
√Tags for data categories remain the same for LC1 and LC2. R, Vc, Vn, Rcl, Spc: see below.
Consider adding tags to completed interpretations to support
secondary analysis (mine and others’).
“Weak tagging” with only case, year, and main subject?
primary: support new analyses.
Are new tags needed to select or exclude published materials?
√Installing plugins and tags for analyzing corpora. LC3 tags Done.
“Page Tagger” plugin enables using tags on pages.
“sk-multi-tag” plugin permits tag-set intersection, but there is no union.
Lacking full set-operations on data items, a tag scheme design to permit foreseeable unions is necessary.
approach: tags more useful for analysis by intersection of data; categories more useful for organizing interpretations; do tags first.
Code Scheme set out in two rows following:
SUBECT: Lng (language); Soc (Socialization); SM (sensory-motor);
Cog (cognitive); Nbr (number); Rdg (reading); Cmp (Computer use)
TIME: W# (week); Q# (quarter); Y# (year)
Vignette names contain day# & subdivision.
RELEVANCE:
Imp (impression); Evl (Evaluation); Int (interpretation); Evd (evidence)
IDEAS: Ntn (notion); Idea (idea); MT (micro-theory); Jdg (Judgment)
DATA: R (all); Vc (video clip); Vn (vignette);
Rcl (recall); Spc (speculation)
What do Tags MEAN, in the current assignment?
If YOU are concerned with such-and-such an issue or period, you may
(or may not) find reference to this vignette useful.
How will the Analyst use tags?
I expect first will come an issue oriented tag-sort through the vignettes by quarter;
after issues come into perspective, this will be followed by a broader, week-by-week review of all the vignettes in the developing time-line to check for other fruitful interactions, as well as the video clips when they are available.
√Work up use-examples in varieties of kinds.
~Testing of “Select and Copy, Paste and Post” process: call it SCPP, (Plan C, see below for B & A)
first collection, fewer than 100 items, succeeded.
second collection, more than 100 but fewer than 200 items cause an “internal server error” at the host.
processing item 3V0620.01 (no. 118); aborted in tags list
now obsolete:
decision: SCCP is a un-usable,
inadequate backstop.
√ develop a real solution.
“Archive-to-Post” process: Plan B (for plan A, see below)
The basic requirement is to take a slice of the database and transfer it into
a form where it can be instantiated as a draft post. What can not be done
“automagically” can be done (with little traffic), by Admin or any contributor.
(See How_to.)
For an automated “email-to-post” process, we will substitute a manual
“select and copy, paste and post” process. Will it work?
Yes, “if you have enough cache you can do anything.”
“Select & Copy, Paste & Post” works.
Link formats lost in copy to post as text. Declare this “bug” a feature, as defining the context of discussion.
But note, every subscriber and visitor will have access in the left sidebar to all tags and any “slice” through the database he wishes to generate and advance for further discussion.
√As a structure development test, replace protocol references in “Sketches of Natural Learning” with links to the protocol source documents. √Decide on value & limits of the practice.
√Extending LC3 tagging to LC1 and LC2
√Extend LC3 tagging to Protocols of LC1
√Done, with modifications below.
√Specify themes in protocol index list.
For LC3 Vignettes, add a leading zero in day date to correct collating sequence error in “tag-slice” archives. √a clerical fix involving many files; simple but time consuming (2.5 days). Done.
√Put in Foreground the Value of Archive for Users and Future √Use Posts LC1, LC2, LC3 as value focused Overviews / Entry points to the database.
√Create examples of Posts for Contributor Demos, e.g., the post “Pure Point” Current draft not compelling.
Plans Concrete Activities: changed to:
√Work up examples of CMS use over summer 2012 √use tags to form Vignette sub-collections
√ design tagging scheme; apply in LC3
√ extend tagging to LC1 & LC2 for integration (at quarterly and annual levels), maintaining separation with LC tags.
Work up examples of CMS use during fall and winter 2012 Extend tagging to inter-relate videos and text materials. Consider carefully video panel names and implications for archive sort order.
Impeded-Plans and Failed-Plans (below)
Digitized Video Analysis & Reformatting on hold until Fall return to USA
Some plug-ins require WP upgrade after US return, e.g.front-side editing Upgrade cautiosly, in fall 2012, or later.
Case Study and Computing chapters require html encoding to bring online.
“Archive-to-Post” process: Plan A: failed on second step.

done: send a message from an NLCSA tag-selected archive to Admin.

current: create a post draft by activation of email received by Admin.

next: encode archives text assembly as email text.
 
complete the “Archive to Post process” for usability by October, 2012
open the text materials to a few interested colleagues by Oct. 2012.
review that experience for evaluation and improvement at 2012 year end.

The current primary effort has failed.
July 7 my hosting agent informed me that the feature on which my current design depends is not supported on my “shared hosting” plan. This is a setback and an impediment. But all is not lost: see Plan B.
The specific failure is non-support by my hosting agent of the standard WordPress email auto-activation and posting process.
Goals
NLCSA publication (online now but hidden from search engines)
Publish comments of “subscribers” after review.
Read and react to “readers’ comments” via email they provide; invite some to be subscribers.,
Enable Contributors to Use Posts for Analysis & Linking Contents
Objectives
NLCSA materials sharing See earlier efforts.
Case Study Content Management System Critical Mass text is well advanced; major effort for video required.
Aims Links to Relevant Published Work
Case Study Materials Sharing LC0bA5 Evolving Datacase Designs NAY

For now, see chapter 2 in book “Case Study and Computing.”

Case Study Method Advancement LC3A1 CASE: Case Analysis Support Environment
LC3cA2 “Soft” Sciences and “Hard” Problems
LC0bA4 Consider the Particular Case and
LC0bA5 (NAY, formatting required.)
Case Study Interpretation See, LC1b NLS; LC2b TIS; and LC3 a post; in the right-sidebar, “On the Method.”
Case Study Archive Construction: NLCSA See LC0a01 With My Heart Upon My Sleeve
Completed Work Links to Relevant Completed Work
Case Study Corpus Construction for LC1, LC2, LC3 Online corpora for LC1, LC2, LC3
Selected Articles (1980-2010) at NLCSA now, see here and here.
Earlier site: “NLCSA.info” Learning and Computing (1998-2004)
Case Study and Computing (w. K. Carley), 1995
Artificial Intelligence and Education, Vol. 1 (Ed. w. M. Yazdani),1987; Vol.2, 1991. See LC0c-AIEd Artificial Intelligence and Education, 1987 & 1991.
Cognition and Computers (w. DuBoulay,Hughes,MacLeod), 1986 See LC1b-NLS Cognition and Computers, 1986.
Computer Experience and Cognitive Development, 1985 See LC2b-TIS Computer Experience and Cognitive Development, 1985
One Child’s Learning (MIT PhD Thesis, Course 6)
Children’s Concept of the English Word (MIT MS Thesis, Course 25)