Monthly Archives: August 2016

Phil 8.31.16

7:00 – 5:00 ASRC

  • Put Wayne’s schedule for October and early November
  • Ping the rest of the Committee
    • Aaron – done
    • Don – done
    • Shimei – done
    • Thom – done
  • Onward with incorporating comments – added ‘fourth estate’ paragraph.
    • I trust my favorite knife because I’ve used it before and I can feel it’s sharpness.
  • Working on building a corpus config file from my GoogleCSE results.
  • Need to add a To Arff menu selection and query.
    • Query is running.
    • Need a binary variable as to whether this is something we want to train on. Probably match plus high quality.

Phil 8.30.16

7:00 – 3:30 ASRC

  • Adding in Wayne’s comments.
  • Got the Corpus generating arff files for BagOfWords and TF-IDF.
  • Here’s the result for NaiveBayes on the first four chapters of Mobey Dick
  • Correctly Classified Instances 3 75 %
    Incorrectly Classified Instances 1 25 %
    Kappa statistic 0.6667
    Mean absolute error 0.125 
    Root mean squared error 0.3536
    Relative absolute error 29.1667 %
    Root relative squared error 71.4435 %
    Total Number of Instances 4 
    
    === Detailed Accuracy By Class ===
    
     TP Rate FP Rate Precision Recall F-Measure MCC ROC Area PRC Area Class
     1.000 0.000 1.000 1.000 1.000 1.000 1.000 1.000 1_-_Loomings
     0.000 0.000 0.000 0.000 0.000 0.000 0.500 0.250 3_-_The_Spouter_Inn
     1.000 0.333 0.500 1.000 0.667 0.577 0.833 0.500 2_-_The_Carpet_Bag
     1.000 0.000 1.000 1.000 1.000 1.000 1.000 1.000 4_-_The_Counterpane
    Weighted Avg. 0.750 0.083 0.625 0.750 0.667 0.644 0.833 0.688 
    
    === Confusion Matrix ===
    
     a b c d <-- classified as
     1 0 0 0 | a = 1_-_Loomings
     0 0 1 0 | b = 3_-_The_Spouter_Inn
     0 0 1 0 | c = 2_-_The_Carpet_Bag
     0 0 0 1 | d = 4_-_The_Counterpane
  • This worked really well: weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4
  • And comparing Jack London stories to Edgar Allen Poe stories works with a corpus of six stories each and not so much with 3 stories each.

Phil 8.29.16

7:00 – 6:00 ASRC

  • Selective Use of News Cues: A Multiple-Motive Perspective on Information Selection in Social Media Environments – Quite close to the Explorer/Confirmer/Avoider study but using a custom(?) browsing interface that tracked the marking of news stories to read later. Subjects were primed for a task with motivations – accuracy, defense and impression. Added this to paragraph 2.9, where explorers are introduced.
  • Looked through Visual Complexity – Mapping Patterns of Information, and it doesn’t even mention navigation. Most information mapping efforts are actually graphing efforts. Added a paragraph in section 2.7
  • Added a TODO for groupthink/confirmation bias, etc.
  • Chat with Heath about AI.He’s looking to build a MUD agent and will probably wind up learning WEKA, etc. so a win, I think.
  • Working on getting the configurator to add string values.
  • Added to DocumentStatistics. Need to switch over to getSourceInfo() from getAddressStrings in the Configurator.
  • Meeting with Wayne about the proposal. One of the branches of conversation went into some research he did on library architecture. That’s been rattling around in my head.
    We tend to talk about interface design where the scale is implicitly for the individual. The environment where these systems function is often thought of as an ecosystem, with the Darwinian perspective that goes along with that. But I think that such a perspective leads to ‘Survival of the Frictionlesss’, where the easiest thing to use wins and damn the larger consequences.
    Reflecting on how the architecture and layout of libraries affected the information interactions of the patrons, I wonder whether we should be thinking about Information Space Architecture. Such a perspective means that the relationships between design at differing scales needs to be considered. In the real world, architecture can encompass everything from the chairs in a room to the landscaping around the building and how that building fits into the skyline.
    I think that regarding information spaces as a designed continuum from the very small to very large is what my dissertation is about at its core. I want a park designed for people, not a wilderness, red in tooth and claw.

Phil 8.26.16

7:00 – 4:00 ASRC

    • Adding more model feedback
    • Something more to think about WRT Group Polarization models? Collective Memory and Spatial Sorting in Animal Groups
    • Need to be able to associate an @attribute  key/value map with Labeled2Dmatrix rows so that we can compare different nominal values across a shared set of numeric columns. This may wind up being a derived class?
      • Working on adding an array of key/value maps;
      • Forgot to add the name to the @data section – oops!
      • text is added to ARFF out. Should I add it to the xlsx outputs as well?
    • Here’s the initial run against the random test data within the class (L2D.arff).
=== Run information ===

Scheme: weka.classifiers.bayes.NaiveBayes
Relation: testdata
Instances: 8
Attributes: 12
name
sv1
sv2
sv3
p1
p2
p3
p4
s1
s2
s3
s4
Test mode: split 66.0% train, remainder test

=== Classifier model (full training set) ===

Naive Bayes Classifier

Class
Attribute p1 p2 p3 p4 s1 s2 s3 s4
(0.13) (0.13) (0.13) (0.13) (0.13) (0.13) (0.13) (0.13)
=======================================================================
sv1
p4-sv1 1.0 1.0 1.0 2.0 1.0 1.0 1.0 1.0
s2-sv1 1.0 1.0 1.0 1.0 1.0 2.0 1.0 1.0
p2-sv1 1.0 2.0 1.0 1.0 1.0 1.0 1.0 1.0
s1-sv1 1.0 1.0 1.0 1.0 2.0 1.0 1.0 1.0
[total] 4.0 5.0 4.0 5.0 5.0 5.0 4.0 4.0

sv2
p2-sv2 1.0 2.0 1.0 1.0 1.0 1.0 1.0 1.0
s4-sv2 1.0 1.0 1.0 1.0 1.0 1.0 1.0 2.0
p1-sv2 2.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
s1-sv2 1.0 1.0 1.0 1.0 2.0 1.0 1.0 1.0
[total] 5.0 5.0 4.0 4.0 5.0 4.0 4.0 5.0

sv3
p2-sv3 1.0 2.0 1.0 1.0 1.0 1.0 1.0 1.0
p1-sv3 2.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
s4-sv3 1.0 1.0 1.0 1.0 1.0 1.0 1.0 2.0
p3-sv3 1.0 1.0 2.0 1.0 1.0 1.0 1.0 1.0
p4-sv3 1.0 1.0 1.0 2.0 1.0 1.0 1.0 1.0
s2-sv3 1.0 1.0 1.0 1.0 1.0 2.0 1.0 1.0
s1-sv3 1.0 1.0 1.0 1.0 2.0 1.0 1.0 1.0
[total] 8.0 8.0 8.0 8.0 8.0 8.0 7.0 8.0

p1
mean 1 0 0 0 1 1 0 0
std. dev. 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667
weight sum 1 1 1 1 1 1 1 1
precision 1 1 1 1 1 1 1 1

p2
mean 0 1 0 0 1 0 1 0
std. dev. 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667
weight sum 1 1 1 1 1 1 1 1
precision 1 1 1 1 1 1 1 1

p3
mean 0 0 1 0 1 0 0 1
std. dev. 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667
weight sum 1 1 1 1 1 1 1 1
precision 1 1 1 1 1 1 1 1

p4
mean 0 0 0 1 1 0 0 1
std. dev. 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667
weight sum 1 1 1 1 1 1 1 1
precision 1 1 1 1 1 1 1 1

s1
mean 1 1 1 1 1 0 0 0
std. dev. 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667
weight sum 1 1 1 1 1 1 1 1
precision 1 1 1 1 1 1 1 1

s2
mean 1 0 0 0 0 1 0 0
std. dev. 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667
weight sum 1 1 1 1 1 1 1 1
precision 1 1 1 1 1 1 1 1

s3
mean 0 1 0 0 0 0 1 0
std. dev. 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667
weight sum 1 1 1 1 1 1 1 1
precision 1 1 1 1 1 1 1 1

s4
mean 0 0 1 1 0 0 0 1
std. dev. 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667 0.1667
weight sum 1 1 1 1 1 1 1 1
precision 1 1 1 1 1 1 1 1



Time taken to build model: 0 seconds

=== Evaluation on test split ===

Time taken to test model on training split: 0 seconds

=== Summary ===

Correctly Classified Instances 0 0 %
Incorrectly Classified Instances 3 100 %
Kappa statistic 0
Mean absolute error 0.2499
Root mean squared error 0.4675
Relative absolute error 108.2972 %
Root relative squared error 133.419 %
Total Number of Instances 3

=== Detailed Accuracy By Class ===

TP Rate FP Rate Precision Recall F-Measure MCC ROC Area PRC Area Class
0.000 0.333 0.000 0.000 0.000 0.000 ? ? p1
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.333 p2
0.000 0.333 0.000 0.000 0.000 0.000 ? ? p3
0.000 0.000 0.000 0.000 0.000 0.000 ? ? p4
0.000 0.000 0.000 0.000 0.000 0.000 0.500 0.500 s1
0.000 0.000 0.000 0.000 0.000 0.000 1.000 1.000 s2
0.000 0.333 0.000 0.000 0.000 0.000 ? ? s3
0.000 0.000 0.000 0.000 0.000 0.000 ? ? s4
Weighted Avg. 0.000 0.000 0.000 0.000 0.000 0.000 0.500 0.611

=== Confusion Matrix ===

a b c d e f g h <-- classified as
0 0 0 0 0 0 0 0 | a = p1
0 0 0 0 0 0 1 0 | b = p2
0 0 0 0 0 0 0 0 | c = p3
0 0 0 0 0 0 0 0 | d = p4
0 0 1 0 0 0 0 0 | e = s1
1 0 0 0 0 0 0 0 | f = s2
0 0 0 0 0 0 0 0 | g = s3
0 0 0 0 0 0 0 0 | h = s4
  • Need to add text data from xml or from other(wrapper info? structured data? UI selections?) sources

Phil 8.25.16

7:00 – 3:30 ASRC

  • Paper
  • Code
    • Build class(s) that uses some of the CorpusBuilder (or just add to output?) codebase to
    • Access webpages based on xml config file
    • Read in, lemmatize , and build bag-of-words per page (configurable max). Done. Took out DF-ITF code and replaced it with BagOfWords in DocumentStatistics.
    • Write out .arff file that includes the following elements
      • @method (TF-IDF, LSI, BOW)
      • @source (loomings, the carpet bag, the spouter inn, the counterpane)
      • @title (Moby-dick, Tarzan)
      • @author (Herman Melville, Edgar Rice Burroughs)
      • @words (nantucket,harpooneer,queequeg,landlord,euroclydon,bedford,lazarus,passenger,circumstance,civilized,water,thousand,about,awful,slowly,supernatural,reality,sensation,sixteen,awake,explain,savage,strand,curbstone,spouter,summer,northern,blackness,embark,tempestuous,expensive,sailor,purse,ocean,tomahawk,black,night,dream,order,follow,education,broad,stand,after,finish,world,money,where,possible,morning,light)
    • So a line should look something like
      • LSI, chapter-1-loomings, Moby-dick, Herman Melville, 0,0,0,0,0,0,0,5,0,0,7,4,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,5,3,3,0,0,0,0,2,0,0,0,5,0,0,3,4,2,0,0,0
      • Updated LabledMatrix2D to generate arff files.

Phil 8.24.16

7:00 – 3:00 ASRC

  • Finished fact checking session
  • Starting contributions. Finished the first pass. That was quick!
  • Ordered my map of the internet

map_of_the_internet_2_0_poster-rcc869f338800460fb9a8086184a4a7e9_w2u_8byvr_512

Phil 8.23.16

7:00 – 4:00 ASRC

  • Continuing to read The Sovereign Map. While thinking about  the Twitter expert paper, I thought that maybe there were mapping projects of the Wikipedia, Schema.org or dmoz.org. I found this for Wikipedia.
  • xkcd maps
  • Paper – continued work on fact-checking/crowdsourced data
  • Code
    • Enable slider in fitnessTest – done
    • Enable reading xml config files – done. Also setting the sliders from load
    • Added Dom4j utils to JavaUtils2
    • get started on WEKA – Starting with Emily’s intro. So far so good! Also ran a Naive Bayes classifier on the weather data set for Aaron to compare.

Phil 8.22.16

7:00 – 2:30 ASRC

Phil 8.19.16

7:00 – 3:30 ASRC

  • Wrote up the action items from the discussion with Thom last night. Now that I have the committee’s initial input, I need to write up an email and re-distribute. Done.
  • Had a thought about the initial GP model. In the fitness test, look for beliefs that are more than ATTRACTION_THRESHOLD similar and be more like them. Possibly look for beliefs that are less than REPULSION_THRESHOLD similar and make the anti-belief more like them. If a statement exists in both belief and antibelief, delete the lowest ranked item, or choose randomly.
    • Working through the logic in beliefMain. I’m just *slow* today.
    • Think I got it. Had to write a method ‘rectifyBeliefs’ that goes in BaseBeliefCA and makes sure that beliefs and antibeliefs don’t overlap. And it’s late enough in the day that I don’t want to try it in the full sim.
  • Working through the fact-checking section
  • Submitted ACM, ICA and UMBC reimbursement requests.

Phil 8.18.16

7:00 – 4:30 ASRC

  • Wrote up my recollection of the meetings with Don, Aaron and Shimei
  • Realized that there should be explicit affordances for confirming and avoiding as well as exploring. Added to the strawman mockup.
  • Adding a ‘model feedback’ section after the evaluation criteria for RQs 2, 3 and 4
  • Back to a simple model of GP
  • Starting on livePrint(String) needs a store and clear method. Done. Also added ctrl-click.
  • Need to make a special test case where one belief chases another. Reworking BeliefMain to support this. Done. The algorithm is good (hacked right now to transform quickly)
  • Meeting with Thom – he wants to see the contributions fleshed out. We discussed the larger frame of the contribution of a self-organizing trustworthy journalism, based on the idea that many people doing small implicit fact checking as a function of their information browsing patterns can be (as?) effective as a few people doing deliberate in-depth fact checking, and with the ability to scale.