No more group meetings after August 6

August 6, 2015

  • Dr. Embley to present HIP paper
  • Dr. Lonsdale gone

July 30, 2015

  • Christopher: switching to ListReader integration; output for third Kilbarchan step isn't clear; still need a way to load the pages from the information given; database-not-updating issue isn't reproducible (hope it was a temporary glitch); noticed that Tae Woo's annotations don't have json field in the database
  • Dr. Woodfield: got hard participation constraints working; not implementing soft participation constraints; working on design refinements; will pack code into jar files eventually, but for now they are linked directly
  • Dr. Embley: working on ListReader integration; found some places where it does run and others where it does not; got back reviews on HIP paper, camera-ready copy due Monday; ran into some strange things – either NetBeans debugger doesn't work right or something else; presenting tomorrow to Robert Kehrer
  • Dr. Lonsdale: made progress on Soar 9.4; it is now committed to the repository; Soar is packaged differently for various platforms, but with same file names; it's going to be very tough to build architecture-dependent jars; needs to make change to the pom.xml file
  • Dr. Liddle: worked with several group members this past week; installed Windows 10; did a fair amount of looking at family history documents
  • Sidebars:
    • ListReader, 3rd cycle and how to load resource pages
    • Participation constraints and functional notation
    • Frontier paper

July 23, 2015

  • Dr. Woodfield: still debugging, making progress; can populate object sets; working on rel sets; “almost done”
  • Christopher: tried to find out why the database wasn't updating; hasn't been able to reproduce the bug; speculates slow network on the day of the problem; looking at ListReader elements to understand in preparation for more ListReader integration; blocker with loading forms and data when a batch gets assigned
  • Dr. Embley: worked more on pipeline; now working on ListReader; met with Thomas and got things going a little bit; worked on JUnit test cases; met with Craig Miller – impressed, want things ready right now; Robert Kehrer wants to meet on the 31st again
  • Dr. Liddle: successes: stake youth conference, kidney stone (A.K.A. Lucifer), girls camp; now back to work
  • Dr. Lonsdale: Soar 9.4 is working; more friendly to later versions of Windows
  • Tae Woo: juggling lots of things; put a button next to the red X, named Regex; wasn't easy to figure out how to do that (in the JavaScript code)
  • Sidebars:
    • Loading forms and data for COMET
    • ListReader test case
    • JUnit issues with ListReader
    • Lonsdale and Liddle, Friday at 10:30am
    • Tae Woo's Regex button

July 16, 2015

  • Dr. Embley: worked on program; generate ground truth for Tae Woo; went over presentation;
  • Dr. Lonsdale: relationship empty variables;
  • Tae Woo: working on getting annotation for regex; working on getting individual records
  • Dr. Woodfield: debugging;
  • Christopher: worked with Tae Woo; spoke with Justin about project; tried to replicate save data problem
  • Sidebars
    • Getting record for Green-FIE

July 9, 2015

  • Tae Woo: was able to run regular expressions in JavaScript with lots of help from Christopher; getting all the information he needs from regex; has been looking at the annotator object to create a regex; has some questions for Christopher
  • Dr. Woodfield: re-read code to do package-level refactoring; added null and unknown objects; looked at how to do general constraints (needs to get something working by September); may want to take a “pragma” approach to putting Java code into general constraints; starting with hard constraints, then soft, then others
  • Christopher: fixed problem with Save JSON button: database sometimes has empty string for JSON field (there's an issue with the Save Data operation – it took three or four saves to get a couple of the examples saved to the database properly); will continue to work on the Save Data bug; ListReader remains a blocker
  • Dr. Embley: Thomas needs to be released from any further obligations; Dr. Embley has the workstation Thomas is using; Thomas will visit Dr. Embley to do a handoff (next Thursday); has Ely page running all the way through nicely; is going back and fixing bugs, edge cases, etc.; recognizes he will have to do some defensive programming; has discovered that somewhere (likely OntoES) the unknowns are set as a space instead of a distinguishing unknown value; ICDAR moved to France
  • Dr. Liddle: worked with several people on aspects of the system; gone next week for stake youth conference
  • Dr. Lonsdale: has batch processing working; installed OntoSoar on Dr. Embley's FamilySearch and BYU machines in both Netbeans and batch mode; OSMX file format changed a little; needs to work on unknowns; has a long list of things to work on
  • Sidebars:
    • Tae Woo and Christopher (as soon as possible)
    • Find single-space “unknown” marker in OntoES

July 2, 2015

  • Dr. Liddle: worked with Christopher on debugging PHP; worked some with Dr. Embley
  • Christopher: Working with Tae Woo on JavaScript (helping Tae Woo learn; Tae Woo is very grateful); has been working on COMET and the new approach for integrating with database and filesystem for the Fe6 ensemble; got JSON pop-up working for Dr. Embley; next up: saving and loading files
  • Dr. Lonsdale: got batch mode running except there is one click in the agent he wants to get rid of in Netbeans; seems to work downstream pretty well
  • Dr. Woodfield: worked on more testing and documentation; will be switching from “industrial code” mode to “fully functional prototype” mode
  • Dr. Embley: finished full (thin) line; had to rewrite all the inference code in Java; works for our Ely page 419 example; started back through with assumptions (like null values, URLs for pages, etc.); i.e. working on the “thick” line now; will bring Thomas' machine into Dr. Embley's office
  • Sidebars:
    • ListReader
    • Programming approach

June 25, 2015

  • Tae Woo: had a job interview and other things; has been studying Stephen Merrill’s code and understands some, but there are some parts that aren’t quite clicking yet for him; would like to meet with Christopher on a regular basis
  • Dr. Woodfield: teaching two classes this term; everything is up and running now; has begun testing his code, doing jUnit tests, but had spurious errors; working now (no more blocker); will be testing for a while
  • Christopher: has been working on models and Ajax for new COMET; has a few questions; please place high priority on User Story #319
  • Dr. Embley: was able to find a recipe for making the merger code to work right (OSMX Wrappers deficiencies); has the merger and enhancer running now; child names is up next
  • Dr. Liddle: working on ER2013 special issue for DKE; spent time helping others in the team with various programming tasks
  • Dr. Lonsdale: for the first time since going to NetBeans, Peter’s code is up and running in it, along with other improvements Dr. Lonsdale has made in the project; will be working on getting it to function in batch mode now (but batch mode won’t necessarily make it easier to run on a non-Windows platform)
  • Sidebars:
    • Dr. Liddle & Christopher — COMET issues — meet Tuesday the 30th at 9:00am
    • Dr. Embley & Dr. Liddle & Christopher — names of directories in our directory structure
    • Dr. Embley & Dr. Liddle — getting to preceding and succeeding pages
    • ListReader email — we need help from Thomas to resolve the ListReader bug identified here

June 18, 2015:

  • Tae Woo: has been trying to learn JavaScript, jQuery, and all that; making progress; working on SQL link next
  • Dr. Embley: got more programming done; has one module ready for refactoring; made some updates to forms merge (pretty well finished); working on standardization of names/places (to do); created an “edited text”, but we may not need it or want it; will probably go to ICDAR on Bill Barrett’s money
  • Dr. Liddle: helped with programming, visited with Christopher and made plans for new COMET; made progress on ER2013 DKE special issue and ER2015 proceedings
  • Dr. Lonsdale: Talked about recent OntoSoar developments at Soar workshop at UMich; top priority for programming is making sure things still work; question: is anything interesting happening with Jiapu project? Answer: yes, creating a competition, also connected Derek with Chinese scholars via George Nagy. With Eric Ringger leaving, more pressure to participate elsewhere in CS.
  • Christopher: worked with Dr. Liddle on COMET database etc.; worked on modifying and adding new PHP models for database
  • Sidebars:
    • end of line hyphen
    • Q2 goals
    • XSD
    • COMET
    • ListReader

June 11, 2015:

  • Christopher: is an uncle! Discovered an issue in COMET – need to deal with PHP stuff first
  • Tae Woo: was looking at Stephen Merrill's JavaScript code; has questions about JavaScript
  • Dr. Embley: submitted HIP paper; may need to travel to Tunisia; split to forms seems to be working; presentation is scheduled for July 17 (w/Craig Miller and others)
  • Dr. Liddle: NSF workshop was quite interesting; lots of things in the stake presidency arena (!)
  • Sidebars:
    • Christopher & Dr. Liddle 9am, Friday, June 12
    • Tae Woo JavaScript questions (jQuery mostly)
    • Christopher & Tae Woo will work on deployment process
    • Dr. Embley & Dr. Liddle: offsets questions
    • Perhaps discuss some COMET things

Schedule Change Starting June 9, move to Thursdays (June 11)

June 2, 2015:

  • Christopher to go directly to Dr. Embley's office
  • Tae Woo to have ordinary individual meeting with Dr. Embley

May 26, 2015:

  • Dr. Lonsdale: got OntoSoar code working; has troubles with PDFIndexer build; needs to get 4-ary relationships going, but has places and other things going
  • Tae Woo: worked on OSMX-to-JSON subsume part; it will merge if it's possible to merge; almost done; looked into JavaScript code on dithers; interested in IDE for JS, but our recommendation is text editor + Chrome Web Developer tools; using annotator.cm
  • Dr. Woodfield: implemented name equivalence we had discussed last week; aside on dates: how do you combine new wine with old bottles? (Okay, I'm interpolating here. ;-) Still needs to extract and populate model instance.
  • Dr. Embley: submitted QMMQ paper; also working on paper with Dr. Nagy and others (submitted); in the middle of writing third paper with Thomas; finished JSON merge to OSMX on the back-end of COMET; it's working for the simple case (one page with no complex annotations or page-crossing annotations); need to do test cases, refactor
  • Christopher: worked with Tae Woo on getting certain info from COMET; got ListReader to take a filename instead of JSON on the command line; seems to work the same as the JSON-on-command-line version; still fails on Christopher's machine, but we think it should run on dithers
  • Dr. Liddle: progress on ER2015, QMMQ paper, worked with Christopher on ListReader interface, worked with Dr. Embley on programming issues, bees
  • Sidebars:
    • Schedule
      • Drs. Woodfield & Lonsdale gone for next two Tuesdays; Dr. Liddle gone next Tuesday; Dr. Embley gone June 4-8; long term; consider changing to Thursdays
    • Names that aren't quite the same but really are the same
    • More discussion on COMET and moving it to the general pipeline
    • Dr. Liddle & Christopher on Friday at 9am; Kilbarchan pages

May 19, 2015:

  • Dr. Liddle: worked with a couple of others on programming issues; ER2015 program decisions have been made! Next up: QMMQ paper and more DKE reviews
  • Tae Woo: made some changes in GreenFIE; created a main class for report feature; added ensemble option as well; looked at JavaScript and did a Hello World for JS; needs specific help on particular topics in JavaScript
  • Dr. Woodfield: did lots of refactoring and working on his model (data structure for holding most info)
  • Christopher: was working on a project and realized it was 5pm, so that explains missing last week; worked with Dr. Liddle on getting ListReader integrated; found command-line parser library; next up: get it into Maven, integrate with ListReader
  • Dr. Embley: spent time coding the form merge to generate OSMX (replacing both David's and Peter's code); spent time writing on ListReader HMM paper; need to get COMET integrated with the directory structure
  • Dr. Lonsdale: got rid of Netbeans issue; now trying to take three development versions of OntoSoar and integrate them into one; need to compare with BookProject; will continue on this
  • Sidebars:
    • Tae Woo and Christopher will get together and discuss JavaScript and so forth
    • Dr. Woodfield's constraint capture in OSMX
    • Christopher & Dr. Liddle will get together to work on Maven integration 9am tomorrow (May 20)

May 12, 2015:

  • Dr. Lonsdale: got a lot of FH stuff done; can't commit code from Family History PC
  • Dr. Liddle: helped various people with various programming tasks; spent time with Christopher on ListReader integration; working on ER2015
  • Peter: has been in Michigan and finished two semesters of classes; thinking and reading about how to do next level of research; will be concentrating on that this summer; wants to start applying it to the family history domain
  • Tae Woo: finished report part of GreenFIE; thinking about how COMET should interact with GreenFIE
  • Dr. Packer: helped Dr. Embley run ListReader on his machine
  • Dr. Woodfield: reviewed a probabilistic cardinality constraints paper for ER2015; has about 1600 lines (clean compile but untested) dealing with model and constraints; first approach is a batch report; has participation constraints and object-set cardinality constraints; also co-occurrence constraints; later: general constraints; tried to generate a few little models and had trouble with relationship-set connections
  • Dr. Embley: updated PowerPoint, created handout, but doesn't have time scheduled yet; revised FROntIER extraction rules and ran it over the full book; has output from both FROntIER and GreenFIE; working around COMET; currently rewriting David's converter for the family history example, hoping to finish this week
  • Sidebars:
    • GreenFIE and COMET integration
    • OSMX blockers for Dr. Woodfield

May 5, 2015:

  • Main topic: Dr. Embley presentation dry run [for Craig Miller & Tom Creighton in mid-May]
  • Prayer: Tae Woo
  • Dr. Liddle: worked with Dr. Embley on ensemble; next up: UI/UX design for ensemble GUI
  • Dr. Packer: debugged ListReader, and it needs testing
  • Tae Woo: was sick; okay now; thinks he's fixed all the bugs in Osmx2Json; needs to write more test cases; works on the current test set; needs to discuss with Dr. Embley
  • Dr. Woodfield: plowed through the paper Liddle gave him
  • Christopher: fixed the problem with deleting the last field in a multiple-field group; fixed the asterisk problem when adding a record above the first one; needs to meet with Dr. Liddle (Thu. 9am)
  • Dr. Embley: finished draft of QMMQ paper, interacted with Joseph to clarify a couple of things; worked on OsmxMerger to decouple from Peter's config file; that works now (lots of refactoring needed); worked on “generated GreenFIE” rules
  • Sidebars:
    • Rename GreenFIE-HD to GreenFIE (after 3pm)

Apr. 28, 2015:

  • Meeting is at 2pm
  • Dr. Liddle: met with Christopher, made progress on ListReader integration; found NPE that needs Dr. Packer's help
  • Dr. Packer: graduated officially; tried running ListReader with new NetBeans looking for problem on Dr. Embley's machine; the problem doesn't replicate on Dr. Packer's machine; will continue looking (more deeply)
  • Christopher: met with Dr. Liddle; not much else to report there; fixed two issues: hover highlighting vs. ListReader highlighting (disabled hover highlighting when in ListReader mode), also issue when populating records from ListReader that is now fixed; might have found the issue with using the “-” key on the last field in a nested group (just needs to find a way to determine the last record)
  • Dr. Woodfield: got a sample OSMX model instance populated; started looking at some high-level design concepts; two design ideas: constraint checkers with plugin architecture (dynamic load from jar file(s)) and constraint handlers that plugin with the constraint checkers
  • Dr. Embley: has finished getting a full line in the agreed-upon directory structure; there is some refactoring that we perhaps should do w.r.t. Peter's merger/evaluator; has created a PowerPoint deck for presentation to be given in a couple of weeks; where does the constraint enforcer fit in, and what are the implications w.r.t. COMET? Biggest thing: discovered that QMMQ paper is due Saturday. Next up: begin to automate the running of tools through the pipeline (as opposed to pushing them through by hand)
  • Sidebars:
    • ListReader NPE
    • OSMX + plugins architecture
    • Liddle/Embley QMMQ and tools quirks
    • Ensemble automation
deg/summer-2015-minutes.txt · Last modified: 2015/07/30 20:32 by swliddle
Back to top
CC Attribution-Share Alike 4.0 International
chimeric.de = chi`s home Valid CSS Driven by DokuWiki do yourself a favour and use a real browser - get firefox!! Recent changes RSS feed Valid XHTML 1.0