Phil 3.22.18

7:00 – ASRC MKT

  • The ONR proposal is in!
  • Promoted the Odyssey thoughts to Phlog
  • More BIC
    • The problem posed by Heads and Tails is not that the players lack a common understanding of salience; it is that game theory lacks an adequate explanation of how salience affects the decisions of rational players. All we gain by adding preplay communication to the model is the realisation that game theory also lacks an adequate explanation of how costless messages affect the decisions of rational players. (pg 180)
  • More TF
    • Invert the ratio for train and validation
    • Add the check against test data
  • Get started on LSTM w/Aaron?

Phil 3.21.18

7:00 – 6:00 ASRC MKT, with some breaks for shovelling

  • First day of spring. Snow on the ground and more in the forecast.
  • I’ve been thinking of ways to describe the differences between information visualizations with respect to maps. Here’s The Odyssey as a geographic map:
  • Odysseus'_Journey
  • The first thing that I notice is just how far Odysseus travelled. That’s about half of the Mediterranean! I thought that it all happened close to Greece. Maps afford this understanding. They are diagrams that support the plotting of trajectories.Which brings me to the point that we lose a lot of information about relationships in narratives. That’s not their point. This doesn’t mean that non-map diagrams don’t help sometimes. Here’s a chart of the characters and their relationships in the Odyssey:
  •  odyssey
  • There is a lot of information here that is helpful. And this I do remember and understood from reading the book. Stories are good about depicting how people interact. But though this chart shows relationships, the layout does not really support navigation. For example, the gods are all related by blood and can pretty much contact each other at will. This chart would have Poseidon accessing Aeolus and  Circe by going through Odysseus.  So this chart is not a map.
  • Lastly, is the relationship that comes at us through search. Because the implicit geographic information about the Odyssey is not specifically in the text, a search request within the corpora cannot produce a result that lets us integrate it
  • OdysseySearchJourney
  • There is a lot of ambiguity in this result, which is similar to other searches that I tried which included travel, sail and other descriptive terms. This doesn’t mean that it’s bad, it just shows how search does not handle context well. It’s not designed to. It’s designed around precision and recall. Context requires a deeper understanding about meaning, and even such recent innovations such as sharded views with cards, single answers, and pro/con results only skim the surface of providing situationally appropriate, meaningful context.
  • Ok, back to tensorflow. Need to update my computer first….
    • Updating python to 64-bit – done
    • Installing Visual Studio – sloooooooooooooooooooooowwwwwwwwwwwww. Done
    • Updating graphics drivers – done
    • Updating tensorflow
    • Updating numpy with intel math
  • At the Validation section in the TF crash course. Good progress. drilling down into all the parts of python that I’ve forgotten. And I got to make a pretty picture: TF_crash_course1

Phil 3.20.18

7:00 – 3:00 ASRC MKT

  • What (satirical) denying a map looks like. Nice application of believability.
  • Need to make a folder with all the CUDA bits and Visual Studio to get all my boxes working with GPU tensorflow
  • Assemble one-page resume for ONR proposal
  • More BIC
    • The fundamental principle of this morality is that what each agent ought to do is to co-operate, with whoever else is co-operating, in the production of the best consequences possible given the behaviour of non-co-operators’ (Regan 1980, p. 124). (pg 167)
    • Ordered On Social Facts
      • Are social groups real in any sense that is independent of the thoughts, actions, and beliefs of the individuals making up the group? Using methods of philosophy to examine such longstanding sociological questions, Margaret Gilbert gives a general characterization of the core phenomena at issue in the domain of human social life.

Back to the TF crash course

    • Had to update my numpy from Christoph Gohlke’s Unofficial Windows Binaries for Python Extension Packages. It’s wonderful, but WHY???
    • Also had this problem updating numpy
      D:\installed>pip3 install "numpy-1.14.2+mkl-cp37-cp37m-win_amd64.whl"
      numpy-1.14.2+mkl-cp37-cp37m-win_amd64.whl is not a supported wheel on this platform.
    • That was solved by installing numpy-1.14.2+mkl-cp36-cp36m-win_amd64.whl. Why cp36 works and cp 37 doesn’t is beyond me.
    • Discussions with Aaron about tasks between now and the TFDS
    • Left early due to snow


Phil 3.19.18

7:00 – 5:00 ASRC MKT

    • The Perfect Selfishness of Mapping Apps
      • Apps like Waze, Google Maps, and Apple Maps may make traffic conditions worse in some areas, new research suggests.
    • Cambridge Social Decision-Making Lab
    • More BIC
      • Schema 3: Team reasoning (from a group viewpoint) pg 153
        • We are the members of S.
        • Each of us identifies with S.
        • Each of us wants the value of U to be maximized.
        • A uniquely maximizes U.
        • Each of us should choose her component of A.
      • Schema 4: Team reasoning (from an individual viewpoint) pg 159
        • I am a member of S.
        • It is common knowledge in S that each member of S identifies
          with S.
        • It is common knowledge in S that each member of S wants the
          value of U to be maximized.
        • It is common knowledge in S that A uniquely maximizes U.
        • I should choose my component of A.
      • Schema 7: Basic team reasoning pg 161
        • I am a member of S.
        • It is common knowledge in S that each member of S identifies
          with S.
        • It is common knowledge in S that each member of S wants the
          value of U to be maximized.
        • It is common knowledge in S that each member of S knows his
          component of the profile that uniquely maximizes U.
        • I should choose my component of the profile that uniquely
          maximizes U.

          • Bacharach notes to himself the ‘hunch’ that this schema is ‘the basic rational capacity’ which leads to high in Hi-Lo, and that it ‘seems to be indispensable if a group is ever to choose the best plan in the most ordinary organizational circumstances’. Notice that Schema 7 does not require that the individual who uses it know everyone’s component of the profile that maximizes U.
      • His hypothesis is that group identification is an individual’s psychological response to the stimulus of a particular decision situation. It is not in itself a group action. (To treat it as a group action would, in Bacharach’ s framework, lead to an infinite regress.) In the theory of circumspect team reasoning, the parameter w is interpreted as a property of a psychological mechanism-the probability that a person who confronts the relevant stimulus will respond by framing the situation as a problem ‘for us’. The idea is that, in coming to frame the situation as a problem ‘for us’, an individual also gains some sense of how likely it is that another individual would frame it in the same way; in this way, the value of w becomes common knowledge among those who use this frame. (Compare the case of the large cube in the game of Large and Small Cubes, discussed in section 4 of the introduction.) Given this model, it seems that the ‘us’ in terms of which the problem is framed must be determined by how the decision situation first appears to each individual. Thus, except in the special case in which w == 1, we must distinguish S (the group with which individuals are liable to identify, given the nature of the decision situation) from T (the set of individuals who in fact identify with S). pg 163
    • Starting with the updates
      C:\WINDOWS\system32>pip3 install --upgrade tensorflow-gpu
      Collecting tensorflow-gpu
        Downloading tensorflow_gpu-1.6.0-cp36-cp36m-win_amd64.whl (85.9MB)
          100% |████████████████████████████████| 85.9MB 17kB/s
      Collecting termcolor>=1.1.0 (from tensorflow-gpu)
        Downloading termcolor-1.1.0.tar.gz
      Collecting absl-py>=0.1.6 (from tensorflow-gpu)
        Downloading absl-py-0.1.11.tar.gz (80kB)
          100% |████████████████████████████████| 81kB 6.1MB/s
      Collecting grpcio>=1.8.6 (from tensorflow-gpu)
        Downloading grpcio-1.10.0-cp36-cp36m-win_amd64.whl (1.3MB)
          100% |████████████████████████████████| 1.3MB 1.1MB/s
      Collecting numpy>=1.13.3 (from tensorflow-gpu)
        Downloading numpy-1.14.2-cp36-none-win_amd64.whl (13.4MB)
          100% |████████████████████████████████| 13.4MB 121kB/s
      Collecting astor>=0.6.0 (from tensorflow-gpu)
        Downloading astor-0.6.2-py2.py3-none-any.whl
      Requirement already up-to-date: six>=1.10.0 in c:\program files\python36\lib\site-packages (from tensorflow-gpu)
      Collecting tensorboard<1.7.0,>=1.6.0 (from tensorflow-gpu)
        Downloading tensorboard-1.6.0-py3-none-any.whl (3.0MB)
          100% |████████████████████████████████| 3.1MB 503kB/s
      Collecting protobuf>=3.4.0 (from tensorflow-gpu)
        Downloading protobuf-3.5.2.post1-cp36-cp36m-win_amd64.whl (958kB)
          100% |████████████████████████████████| 962kB 1.3MB/s
      Collecting gast>=0.2.0 (from tensorflow-gpu)
        Downloading gast-0.2.0.tar.gz
      Requirement already up-to-date: wheel>=0.26 in c:\program files\python36\lib\site-packages (from tensorflow-gpu)
      Requirement already up-to-date: html5lib==0.9999999 in c:\program files\python36\lib\site-packages (from tensorboard<1.7.0,>=1.6.0->tensorflow-gpu)
      Requirement already up-to-date: bleach==1.5.0 in c:\program files\python36\lib\site-packages (from tensorboard<1.7.0,>=1.6.0->tensorflow-gpu)
      Requirement already up-to-date: markdown>=2.6.8 in c:\program files\python36\lib\site-packages (from tensorboard<1.7.0,>=1.6.0->tensorflow-gpu)
      Requirement already up-to-date: werkzeug>=0.11.10 in c:\program files\python36\lib\site-packages (from tensorboard<1.7.0,>=1.6.0->tensorflow-gpu)
      Collecting setuptools (from protobuf>=3.4.0->tensorflow-gpu)
        Downloading setuptools-39.0.1-py2.py3-none-any.whl (569kB)
          100% |████████████████████████████████| 573kB 2.3MB/s
      Building wheels for collected packages: termcolor, absl-py, gast
        Running bdist_wheel for termcolor ... done
        Stored in directory: C:\Users\philip.feldman\AppData\Local\pip\Cache\wheels\de\f7\bf\1bcac7bf30549e6a4957382e2ecab04c88e513117207067b03
        Running bdist_wheel for absl-py ... done
        Stored in directory: C:\Users\philip.feldman\AppData\Local\pip\Cache\wheels\3c\0f\0a\6c94612a8c26070755559045612ca3645fea91c11f2148363e
        Running bdist_wheel for gast ... done
        Stored in directory: C:\Users\philip.feldman\AppData\Local\pip\Cache\wheels\8e\fa\d6\77dd17d18ea23fd7b860e02623d27c1be451521af40dd4a13e
      Successfully built termcolor absl-py gast
      Installing collected packages: termcolor, absl-py, setuptools, protobuf, grpcio, numpy, astor, tensorboard, gast, tensorflow-gpu
        Found existing installation: setuptools 38.4.0
          Uninstalling setuptools-38.4.0:
            Successfully uninstalled setuptools-38.4.0
        Found existing installation: protobuf 3.5.1
          Uninstalling protobuf-3.5.1:
            Successfully uninstalled protobuf-3.5.1
        Found existing installation: numpy 1.13.0+mkl
          Uninstalling numpy-1.13.0+mkl:
            Successfully uninstalled numpy-1.13.0+mkl
        Found existing installation: tensorflow-gpu 1.4.0
          Uninstalling tensorflow-gpu-1.4.0:
            Successfully uninstalled tensorflow-gpu-1.4.0
      Successfully installed absl-py-0.1.11 astor-0.6.2 gast-0.2.0 grpcio-1.10.0 numpy-1.14.2 protobuf-3.5.2.post1 setuptools-39.0.1 tensorboard-1.6.0 tensorflow-gpu-1.6.0 termcolor-1.1.0
    • That caused the following items to break when I tried running “”
      "C:\Program Files\Python36\python.exe" D:/Development/Sandboxes/TensorflowPlayground/HelloPackage/
      Traceback (most recent call last):
        File "C:\Program Files\Python36\lib\site-packages\tensorflow\python\platform\", line 75, in preload_check
        File "C:\Program Files\Python36\lib\ctypes\", line 348, in __init__
          self._handle = _dlopen(self._name, mode)
      OSError: [WinError 126] The specified module could not be found
      During handling of the above exception, another exception occurred:
      Traceback (most recent call last):
        File "D:/Development/Sandboxes/TensorflowPlayground/HelloPackage/", line 28, in 
          import tensorflow as tf
        File "C:\Program Files\Python36\lib\site-packages\tensorflow\", line 24, in 
          from tensorflow.python import *
        File "C:\Program Files\Python36\lib\site-packages\tensorflow\python\", line 49, in 
          from tensorflow.python import pywrap_tensorflow
        File "C:\Program Files\Python36\lib\site-packages\tensorflow\python\", line 30, in 
        File "C:\Program Files\Python36\lib\site-packages\tensorflow\python\platform\", line 82, in preload_check
          % (build_info.cudart_dll_name, build_info.cuda_version_number))
      ImportError: Could not find 'cudart64_90.dll'. TensorFlow requires that this DLL be installed in a directory that is named in your %PATH% environment variable. Download and install CUDA 9.0 from this URL:
    • Installing Visual Studio for the DLLs before I install the Cuda parts
    • Downloading cuda_9.0.176_win10.exe from here There are also two patches
    • Next set of errors
      Traceback (most recent call last):
        File "D:/Development/Sandboxes/TensorflowPlayground/HelloPackage/", line 28, in 
          import tensorflow as tf
        File "C:\Program Files\Python36\lib\site-packages\tensorflow\", line 24, in 
          from tensorflow.python import *
        File "C:\Program Files\Python36\lib\site-packages\tensorflow\python\", line 49, in 
          from tensorflow.python import pywrap_tensorflow
        File "C:\Program Files\Python36\lib\site-packages\tensorflow\python\", line 30, in 
        File "C:\Program Files\Python36\lib\site-packages\tensorflow\python\platform\", line 97, in preload_check
          % (build_info.cudnn_dll_name, build_info.cudnn_version_number))
      ImportError: Could not find 'cudnn64_7.dll'. TensorFlow requires that this DLL be installed in a directory that is named in your %PATH% environment variable. Note that installing cuDNN is a separate step from installing CUDA, and this DLL is often found in a different directory from the CUDA DLLs. You may install the necessary DLL by downloading cuDNN 7 from this URL:
  • Looking for cudnn64_7.dll here?
  • Aaaand that seems to be working!
  • Tweaked ONR proposal with Aaron. Discovered that there is one page per PI, so we need to make one-page resumes.



Phil 3.17.18

This came to me as a herding paper title for SASO 2018: This one simple trick can disrupt digital societies. Too much?

First International workshop on Socio-Cognitive Systems

  •  In this workshop we want to explore the interactions between cognitive and social aspects of “socio-cognitive systems” – that is where the social and cognitive aspects are studied together. The workshop connects elements of IJCAI/ECAI, AAMAS and ICML. Of course, modelling these systems in terms of Multi-Agent Systems seems intuitive, but also would require special attention to the social concepts in these MAS. The cognitive abilities of the agents should adapt themselves to the social context and development, which connects this area to machine learning in a social context. 
  • Deadline for submissions: 1st May 2018
  • Notification of acceptance: May/June 2018
  • Camera-ready copy of papers: June 2018
  • Workshop: 14 or 15 July 2018
  • Stockholm, co-located with IJCAI

Symmetric generative methods and t-SNE: a short survey

  • In data visualization, a family of methods is dedicated to the symmetric numerical matrices which contain the distances or similarities between high-dimensional data vectors. The method t-Distributed Stochastic Neighbor Embedding has been recently introduced for data visualization. Leading to competitive nonlinear embeddings which are able to reveal the natural classes, several variants have been developed. For comparison purposes, it is presented the recent generative alternative methods (Glove, probabilistic CA, LSPM, LargeVis, SBM) in the literature for nonlinear embedding via low dimensional positions.


  • étudier is a small Python program that uses Selenium and requests-html to drive a non-headless browser to collect a citation graph around a particular Google Scholar citation. The resulting network is written out as a Gephi file and a D3 visualization using networkx.

Phil 3.16.18

7:00 – 4:00 ASRC MKT

    • Umwelt
      • In the semiotic theories of Jakob von Uexküll and Thomas A. Sebeokumwelt (plural: umwelten; from the German Umwelt meaning “environment” or “surroundings”) is the “biological foundations that lie at the very epicenter of the study of both communication and signification in the human [and non-human] animal”.[1] The term is usually translated as “self-centered world”.[2] Uexküll theorised that organisms can have different umwelten, even though they share the same environment. The subject of umwelt and Uexküll’s work is described by Dorion Sagan in an introduction to a collection of translations.[3] The term umwelt, together with companion terms umgebungand innenwelt, have special relevance for cognitive philosophers, roboticists and cyberneticians, since they offer a solution to the conundrum of the infinite regress of the Cartesian Theater.
    • Benjamin Kuipers
      • How Can We Trust a Robot? (video)
        • Advances in artificial intelligence (AI) and robotics have raised concerns about the impact on our society of intelligent robots, unconstrained by morality or ethics
      • Socially-Aware Navigation Using Topological Maps and Social Norm Learning
        • We present socially-aware navigation for an intelligent robot wheelchair in an environment with many pedestrians. The robot learns social norms by observing the behaviors of human pedestrians, interpreting detected biases as social norms, and incorporating those norms into its motion planning. We compare our socially-aware motion planner with a baseline motion planner that produces safe, collision-free motion. The ability of our robot to learn generalizable social norms depends on our use of a topological map abstraction, so that a practical number of observations can allow learning of a social norm applicable in a wide variety of circumstances. We show that the robot can detect biases in observed human behavior that support learning the social norm of driving on the right. Furthermore, we show that when the robot follows these social norms, its behavior influences the behavior of pedestrians around it, increasing their adherence to the same norms. We conjecture that the legibility of the robot’s normative behavior improves human pedestrians’ ability to predict the robot’s future behavior, making them more likely to follow the same norm.
    • Erin’s defense
      • Nice slides!
      • Slide 4 – narrowing from big question to dissertation topic. Nice way to set up framing
      • Intellectual function vs. adaptive behavior
      • Loss of self-determination
      • Maker culture as a way of having your own high-dimensional vector? Does this mean that the maker culture is inherently more exploratory when compared to …?
      • “Frustration is an easy way to end up in off-task behavior”
      • Peer learning as gradient descent?
      • Emic ethnography
      • Pervasive technology in education
      • Turn-taking
      • Antecedent behavior consequence theory
      • Reducing the burden on the educators. Low-level detection and to draw attention to the educator and annotate. Capturing and labeling
      • Helina – bring the conclusions back to the core questions
      • Diversity injection works! Mainstream students gained broader appreciation of students with disability
      • Q: Does it make more sense to focus on potentially charismatic technologies that will include the more difficult outliers even if it requires a breakthrough? Or to make incremental improvements that can improve accessibility to some people with disabilities faster?
      • Boris analytic software


Phil 3.15.18

8:30 – 4:30 ASRC MKT

Phil 3.14.18

7:00 – 4:00 ASRC MKT

  • Cannot log into my timesheet
  • Continuing along with TF. Got past the introductions and to the beginning of the coding.
  • Myanmar: UN blames Facebook for spreading hatred of Rohingya (The Guardia)
    • ‘Facebook has now turned into a beast’, says United Nations investigator, calling network a vehicle for ‘acrimony, dissension and conflict’
  • Related to the above (which was pointed out by the author in this tweet)
  • Keynote: Susan Dumais
    • Better Together: An Interdisciplinary Perspective on Information Retreival
    • A solution to plato’s problem – latent semantic indexing
    • The road to LSI
    • LSI paper as dimension reduction Dumas et al 1988,
    • Search and context
      • Ranked list of 10 blue links
      • Need to understand the context in which they occur. Documents are intricately linked
      • Search is doe to accomplish something (picture of 2 people pointing at a chart/map?)
      • Short and long term models of interest (Bennett et al 2012)
      • Stuff I’ve Seen (2003) Becomes LifeBrowser
    • Future directions
      • ML will take over IR for better or worst
      • Moving from a world that indexe strings to a world that indexes things
      • Bing is doing pro/con with questions, state maintained dialog
  • Here and Now: Reality-Based Information Retrieval. [Perspective Paper]
    Wolfgang Büschel, Annett Mitschick and Raimund Dachselt

    • Perspective presentation on AR-style information retreival.
    • Maybe an virtual butler that behaves like an invisible freind?
  • A Study of Immediate Requery Behavior in Search.
    Haotian Zhang, Mustafa Abualsaud and Mark Smucker
  • Exploring Document Retrieval Features Associated with Improved Short- and Long-term Vocabulary Learning Outcomes.
    Rohail Syed and Kevyn Collins-Thompson
  • Switching Languages in Online Searching: A Qualitative Study of Web Users’ Code-Switching Search Behaviors.
    Jieyu Wang and Anita Komlodi
  • A Comparative User Study of Interactive Multilingual Search Interfaces.
    Chenjun Ling, Ben Steichen and Alexander Choulos

Phil 3.13.18

7:00 – 5:00 ASRC MKT

  • Sent T a travel request for the conference. Yeah, it’s about as late as it could be, but I just found out that I hadn’t registered completely…
  • Got Tensorflow running on my laptop. Can’t get Python 2.x warnings to not show. Grrrr.
  • Had to turn off privacy badger to get the TF videos to play. Nicely done
  • Information Fostering – Being Proactive with Information Seeking and Retrieval [Perspective Paper]
    Chirag Shah

    • Understanding topic, task, and intention
    • People are boxed in when looking for information. Difficult to encouraging broad thinking
    • Ryan White – tasks? Cortana?
    • What to do when things go bad:
  • The Role of the Task Topic in Web Search of Different Task Types.
    Daniel Hienert, Matthew Mitsui, Philipp Mayr, Chirag Shah and Nicholas Belkin
  • Juggling with Information Sources, Task Type, and Information Quality
    Yiwei Wang, Shawon Sarkar and Chirag Shah

    • Doing tasks in a study has an odd bias that drives users to non-social information sources. Since the user is not engaged in a “genuine” task, the request of other people isn’t considered as viable.
  • ReQuIK: Facilitating Information Discovery for Children Through Query Suggestions.
    Ion Madrazo, Oghenemaro Anuyah, Nevena Dragovic and Maria Soledad Pera

    • LSTM model + hand-coded heuristics combined deep and wide. LSTM produces 92% accuracy, Hand-rolled 68%, both 94%
    • Wordnet-based similarity
  • Improving exploration of topic hierarchies: comparative testing of simplified Library of Congress Subject Heading structures.
    Jesse David Dinneen, Banafsheh Asadi, Ilja Frissen, Fei Shu and Charles-Antoine Julien

    • Pruning large scale structures to support visualization
    • Browsing complexity calculations
    • Really nice. Dynamically pruned trees, with the technical capability for zooming at a local level
  • Fixation and Confusion – Investigating Eye-tracking Participants’ Exposure to Information in Personas.
    Joni Salminen, Jisun An, Soon-Gyo Jung, Lene Nielsen, Haewoon Kwak and Bernard J. Jansen

    • LDA topic extraction
    • Eyetribe – under $200. Bought by Facebook
    • Attribute similarity as a form of diversity injection
  • “I just scroll through my stuff until I find it or give up”: A Contextual Inquiry of PIM on Private Handheld Devices.
    Amalie Jensen, Caroline Jægerfelt, Sanne Francis, Birger Larsen and Toine Bogers

    • contextual inquiry – good at uncovering tacit interactions
    • Looking at the artifacts of PIM
  • Augmentation of Human Memory: Anticipating Topics that Continue in the Next Meeting
    Seyed Ali Bahrainian and Fabio Crestani

    • Social Interactions Log Analysis System (Bahrainian et. al)
    • Proactive augmentation of memory
    • LDA topic extraction
    • Recency effect could apply to distal ends of a JuryRoom discussion
  • Characterizing Search Behavior in Productivity Software.
    Horatiu Bota, Adam Fourney, Susan Dumais, Tomasz L. Religa and Robert Rounthwaite

Phil 3.12.18

7:00 – 7:00 ASRC

  • The Surprising Creativity of Digital Evolution: A Collection of Anecdotes from the Evolutionary Computation and Artificial Life Research Communities
    • Biological evolution provides a creative fount of complex and subtle adaptations, often surprising the scientists who discover them. However, because evolution is an algorithmic process that transcends the substrate in which it occurs, evolution’s creativity is not limited to nature. Indeed, many researchers in the field of digital evolution have observed their evolving algorithms and organisms subverting their intentions, exposing unrecognized bugs in their code, producing unexpected adaptations, or exhibiting outcomes uncannily convergent with ones in nature.
  • Analyzing Knowledge Gain of Users in Informational Search Sessions on the Web.
    Ujwal Gadiraju, Ran Yu, Stefan Dietze and Peter Holtz
  • Query Priming for Promoting Critical Thinking in Web Search.
    Yusuke Yamamoto and Takehiro Yamamoto

    • TruthFinder – consistency
    • CowSearch – provides supporting information for credibility judgements
    • Query priming only worked on university-educated participants. Explorer? Or not university educated are stampede?
  • Searching as Learning: Exploring Search Behavior and Learning Outcomes in Learning-related Tasks.
    Souvick Ghosh, Manasa Rath and Chirag Shah

    • Structures of the Life-World
    • Distinguish, organize and conclude are commonly used words by participants describing their tasks. This implies that learning, or at least the participant’s view of learning is building an inventory of facts. Hmm.
    • Emotional effect on cognitive behavior? It would be interesting to see if (particularly with hot-button issues), the emotion can lead to a more predictable dimension reduction.
  • Informing the Design of Spoken Conversational Search [Perspective Paper]
    Johanne R Trippas, Damiano Spina, Lawrence Cavedon, Hideo Joho and Mark Sanderson

    •  Mention to Johanne about spoken interface to SQL
    • EchoQuery
  • Style and alignment in information-seeking conversation.
    Paul Thomas, Mary Czerwinski, Daniel Mcduff, Nick Craswell and Gloria Mark

    • Conversational Style (Deborah Tannen) High involvement and High consideration.
    • Alignment. Match each others patterns of speech!
    • Joint action, interactive alignment, and dialog
      • Dialog is a joint action at different levels. At the highest level, the goal of interlocutors is to align their mental representations. This emerges from joint activity at lower levels, both concerned with linguistic decisions (e.g., choice of words) and nonlinguistic processes (e.g., alignment of posture or speech rate). Because of the high-level goal, the interlocutors are particularly concerned with close coupling at these lower levels. As we illustrate with examples, this means that imitation and entrainment are particularly pronounced during interactive communication. We then argue that the mechanisms underlying such processes involve covert imitation of interlocutors’ communicative behavior, leading to emulation of their expected behavior. In other words, communication provides a very good example of predictive emulation, in a way that leads to successful joint activity.
  • SearchBots: User Engagement with ChatBots during Collaborative Search.
    Sandeep Avula, Gordon Chadwick, Jaime Arguello and Robert Capra

Phil 3.11.18

7:00 – 5:00 ASRC MKT

  • Notes from Coursera Deep Learning courses by Andrew Ng. Cool notes by Tess Ferrandez <- nice Angular stuff here too
  • Kill Math project for math visualizations
    • The power to understand and predict the quantities of the world should not be restricted to those with a freakish knack for manipulating abstract symbols.
  • Leif Azzopardi
  • CHIIR 2018 DC today! I’m on after lunch! Impostor syndrome well spun up right now
    • Contextualizing Information Needs of Patients with Chronic Conditions Using Smartphones
      • Henna Kim
      • What about the OpenAPS project?
      • recognition that patients need pieces of information to accomplish health related work to better manage their condition to health and wellness
      • Information needs arise from talks???
      • Goals that patients pursue for a long period of time
  • Task-based Information Seeking in Different Study Settings
    • Yiwei Wang
    • People are influenced by their natural environment. Also the cognitive environment
    • What about nomadic/flock/stampede?
    • She needs a research browser!
    • Need for cognition
  • The Moderator Effect of Working Memory and Emotion on the Relationship between Information Overload and Online Health Information Quality
    • Yung-Sheng Chang
    • Information overload and information behavior/attitude
    • Overload is also the inability to simplify. Framing should help with incorporation
  • Exploring the effects of social contexts on task-based information seeking behavior
    • Eun Youp Rha
    • Socio-cultural context
    • A task is only recognizable within a certain context when people agree it is a task
    • Sociocultural mental processes. Perception, memory, Classification signification (Zerubavel, 1997)
      • Sociology of perception
      • Sociology of attention
      • Practice theory – Viewing human actions as regular performances of ritualized actions
    • How do tow communities in different places evolve different norms?
  • Distant Voices in the Dark: Understanding the incongruent information needs of fiction authors and readers
    • Carol Butler
    • Authors and readers interact with each other
    • What about The Martian?
    • Also, fanfiction?
    • Authors want to interact with other authors, readers with readers.
    • Also writing for peers where readers are assumed not to exist (technical publications)
    • Writing and reading is built around an industrial process (mass entertainment in general? What about theater?)
    • Stigma around self-publishing
    • Not much need to interact because they don’t get that much from each other. Also, the book has just been released and the readers haven’t read it. What question do you ask when you haven’t read the book yet? This leads to the “same stupid questions”
    • Library catalogs that incorporate social media. Sense is that it failed?
    • BookTube?
  • On the Interplay Between Search Behavior and Collections in Digital Libraries and Archives
    • Tessel Bogaard
    • Digital library, with text, meta information, clickstreams in logs
    • How do we let the domain curators understand their users
    • Family announcements are disproportionately popular. Short sessions, with few clicks and documents
    • WWII documents are from prolonged interactions
    • Grouping sessions using k medoid using user interactions  and facets. Use average silhouette widths (how similar are the clusters) Stability over time
    • Markov cahin analysis
    • Side by side comparison over teh whole data set
    • Session graph (published demo paper)
  • Creative Search: Using Search to Leverage Your Everyday Creativity
    • Yinglong Zhang
    • Creativity can be taught
    • To be creative, you need to acquire deep domain knowledge. High dimensions. Implies that thinking in low dimensions are creativity constraining.
    • Crowdsourcing tools (Yu, Kittur, and Kraut 2016)
    • Free form web curation (Kerne et. al)
  • Diversity-Enhanced Recommendation Interface and Evaluation
    • Chun-Hua Tsai
    • Diversity-enhanced interface design
    • Continuous Controlability and experience
    • Very LMN-like
    • Interface is swamped by familiarity. Minimum delta from current interfaces.
  • Towards Human-Like Conversational Search Systems
    • Mateusz Dubiel
    • More experience = more use.
    • Needs more conversational?
    • Enable navigation through converation?
    • Back chaining and forward chaining
    • Asking for clarification
    • Turn taking
  • Room 225
  • Journal of information research
  • Paul Thomas (MS Research)
  • Ryan White (MS Research)
  • Jimmy Lin (Ex Twitter)
  • Dianne Kelly.

Phil 3.9.18

8:00 – 4:30 ASRC MKT

  • Still working on the nomad->flocking->stampede slide. Do I need a “dimensions” arrow?
  • Labeled slides. Need to do timings – done
  • And then Aaron showed up, so lots of reworking. Done again!
  • Put the ONR proposal back in its original form
  • An overview of gradient descent optimization algorithm
    • Gradient descent is one of the most popular algorithms to perform optimization and by far the most common way to optimize neural networks. At the same time, every state-of-the-art Deep Learning library contains implementations of various algorithms to optimize gradient descent (e.g. lasagne’scaffe’s, and keras’ documentation). These algorithms, however, are often used as black-box optimizers, as practical explanations of their strengths and weaknesses are hard to come by. This blog post aims at providing you with intuitions towards the behaviour of different algorithms for optimizing gradient descent that will help you put them to use.

Phil 3.8.18

7:00 – 5:00 ASRC

  • Another nice comment from Joanna Bryson on BBC Business Daily – The bias is seldom in the algorithm. Latent Semantic Indexing is simple arithmetic. The data contains the bias, and that’s from us. Fairness is a negotiated concept, which means that is is complicated. Requiring algorithmic fairness necessitates placing enormous power in the hands of those writing the algorithms.
  • The science of fake news (Science magazine)
    • The rise of fake news highlights the erosion of long-standing institutional bulwarks against misinformation in the internet age. Concern over the problem is global. However, much remains unknown regarding the vulnerabilities of individuals, institutions, and society to manipulations by malicious actors. A new system of safeguards is needed. Below, we discuss extant social and computer science research regarding belief in fake news and the mechanisms by which it spreads. Fake news has a long history, but we focus on unanswered scientific questions raised by the proliferation of its most recent, politically oriented incarnation. Beyond selected references in the text, suggested further reading can be found in the supplementary materials.
  • Incorporating Sy’s comments into a new slide deck
  • More ONR
  • Meeting with Shimei
    • Definitely use the ONR-specified headings
    • Research is looking good and interesting! Had to spend quite a while explaining lexical trajectories.
  • Ran through the slides with Sy again. Mostly finalized?

Phil 3.7.18

7:00 – 5:00 ASRC MKT

  • Some surprising snow
  • Meeting with Sy at 1:30 slides
  • Meeting with Dr. DesJardins at 4:00
  • Nice chat with Wajanat about the presentation of the Saudi Female self in physical and virtual environments
  • Sprint planning
    • Finish ONR Proposal VP-331
    • CHIIR VP-332
    • Prep for TF dev conf VP-334
    • TF dev conf VP-334
  • Working on the ONR proposal
  • Oxford Internet Institute – Computational Propaganda Research Project
    • The Computational Propaganda Research Project (COMPROP) investigates the interaction of algorithms, automation and politics. This work includes analysis of how tools like social media bots are used to manipulate public opinion by amplifying or repressing political content, disinformation, hate speech, and junk news. We use perspectives from organizational sociology, human computer interaction, communication, information science, and political science to interpret and analyze the evidence we are gathering. Our project is based at the Oxford Internet Institute, University of Oxford.
    • Polarization, Partisanship and Junk News Consumption over Social Media in the US
      • What kinds of social media users read junk news? We examine the distribution of the most significant sources of junk news in the three months before President Donald Trump’s first State of the Union Address. Drawing on a list of sources that consistently publish political news and information that is extremist, sensationalist, conspiratorial, masked commentary, fake news and other forms of junk news, we find that the distribution of such content is unevenly spread across the ideological spectrum. We demonstrate that (1) on Twitter, a network of Trump supporters shares the widest range of known junk news sources and circulates more junk news than all the other groups put together; (2) on Facebook, extreme hard right pages—distinct from Republican pages—share the widest range of known junk news sources and circulate more junk news than all the other audiences put together; (3) on average, the audiences for junk news on Twitter share a wider range of known junk news sources than audiences on Facebook’s public pages
      • Need to look at the variance in the articles. Are these topical stampedes? Or is this source-oriented?
  • Understanding and Addressing the Disinformation Ecosystem
    • This workshop brings together academics, journalists, fact-checkers, technologists, and funders to better understand the challenges produced by the current disinformation ecosystem. The facilitated discussions will highlight relevant research, share best-practices, identify key questions of scholarly and practical concern regarding the nature and implications of the disinformation ecosystem, and outline a potential research agenda designed to answer these questions.
  • More BIC
    • The psychology of group identity allows us to understand that group identification can be due to factors that have nothing to do with the individual preferences. Strong interdependence and other forms of common individual interest are one sort of favouring condition, but there are many others, such as comembership of some existing social group, sharing a birthday, and the artificial categories of the minimal group paradigm. (pg 150)
    • Wherever we may expect group identity we may also expect team reasoning. The effect of team reasoning on behavior is different from that of individualistic reasoning. We have already seen this for Hi-Lo. This has wide implications. It makes the theory of team reasoning a much more powerful explanatory and predictive theory than it would be if it came on line only in games with th3e right kind of common interest. To take just one example, if management brings it about so that the firm’s employees identify with the firm, we may expect for them to team-reason and so to make choices that are not predicted by the standard theories of rational choice. (pg 150)
    • As we have seen, the same person passes through many group identities in the flux of life, and even on a single occasion more than one of these identities may be stimulated. So we will need a model of identity in which the probability of a person’s identification is distributed over not just two alternatives-personal self-identity or identity with a fixed group-but, in principle, arbitrarily many. (pg 151)

Phil 3.6.18

7:00 – 4:00 ASRC MKT

  • Endless tweaking of the presentation
    • Pinged Sy – Looks like something on Wednesday. Yep his place around 1:30
  • More BIC
    • The explanatory potential of team reasoning is not confined to pure coordination games like Hi-Lo. Team reasoning is assuredly important for its role in explaining the mystery facts about Hi-Lo; but I think we have stumbled on something bigger than a new theory of behaviour in pure coordination games. The key to endogenous group identification is not identity of interest but common interest giving rise to strong interdependence. There is common interest in Stag Hunts, Battles of the Sexes, bargaining games and even Prisoner’s Dilemmas. Indeed, in any interaction modelable as a ‘mixed motive’ game there is an element of common interest. Moreover, in most of the landmark cases, including the Prisoner’s Dilemma, the common interest is of the kind that creates strong interdependence, and so on the account of chapter 2 creates pressure for group identification. And given group identification, we should expect team reasoning. (pg 144)
    • There is a second evolutionary argument in favour of the spontaneous team-reasoning hypothesis. Suppose there are two alternative mental mechanisms that, given common interest, would lead humans to act to further that interest. Other things being equal, the cognitively cheapest reliable mechanism will be favoured by selection. As Sober and Wilson (1998) put it, mechanisms will be selected that score well on availability, reliability and energy efficiency. Team reasoning meets these criteria; more exactly, it does better on them than the alternative heuristics suggested in the game theory and psychology literature for the efficient solution of common-interest games. (pg 146)
    • BIC_pg 149 (pg 149)
  • Educational resources from machine learning experts at Google
    • We’re working to make AI accessible by providing lessons, tutorials and hands-on exercises for people at all experience levels. Filter the resources below to start learning, building and problem-solving.
  • A Structured Response to Misinformation: Defining and Annotating Credibility Indicators in News Articles
    • The proliferation of misinformation in online news and its amplification by platforms are a growing concern, leading to numerous efforts to improve the detection of and response to misinformation. Given the variety of approaches, collective agreement on the indicators that signify credible content could allow for greater collaboration and data-sharing across initiatives. In this paper, we present an initial set of indicators for article credibility defined by a diverse coalition of experts. These indicators originate from both within an article’s text as well as from external sources or article metadata. As a proof-of-concept, we present a dataset of 40 articles of varying credibility annotated with our indicators by 6 trained annotators using specialized platforms. We discuss future steps including expanding annotation, broadening the set of indicators, and considering their use by platforms and the public, towards the development of interoperable standards for content credibility.
    • Slide deck for above
  • Sprint review
    • Presented on Talk, CI2018 paper, JuryRoom, and ONR proposal.
  • ONR proposal
    • Send annotated copy to Wayne, along with the current draft. Basic question is “is this how it should look? Done
    • Ask folks at school for format help?