Archive for November, 2014

Weekly QuEST Discussion Topics, 21 Nov

November 20, 2014 Leave a comment

The first QuEST meeting of any calendar year we dedicate to the Kabrisky Memorial Lecture.  One of the original founding members of our group suggested early on the first meeting of each calendar year we give a fresh over view of “What is QuEST.”  This year I want to spend our remaining meetings at then end of this calendar year to review some of the material we’ve hit throughout the year for consideration for inclusion in our overview Kabrisky lecture.  Keep in mind the goal of the Kabrisky lecture is to have a single overview talk that doesn’t explain all the details of our material but provides the insights for someone from the outside can get an understanding of our effort without any background in our jargon.  So I invite everyone to pick through your notes / the blog and bring up things you want us to consider as important points that should be included in the “What is QuEST – Kabrisky Memorial Lecture”.  Specifically we want to ensure the overview talk captures most of the ideas of QuEST maybe without all the details.  By my count there appears to be 4 QuEST meetings left before the Kabrisky lecture on the 9th of Jan.  So this week we will review topics from Jan-Mar 2014.

As an example of the types of discussions I want to have in these meetings till the end of the year:

Started the year considering LaRue and specifically how to do tenet consistent Type 2 processes using that architecture. – we ended the year considering how to do the same thing using a combination of ACT-r and Semantic networks infrastructure / ontology software and conventional machine learning approaches for the Mike R. effort.  So we want to start this week with a discussion on what are our current positions on implementing dual process solutions using existing software / frameworks… – so the question is how should we position our interest in existing cognitive models in the overview talk?

Categories: Uncategorized

Weekly QuEST Discussion Topics and News, 14 Nov

November 13, 2014 Leave a comment

This week, our QuEST colleague Steven Harbour will present his research and lead a discussion on:

A Neuroergonomic Quasi-Experiment: Predictors of Situation Awareness
and Display Usability with USAF Pilots while Performing Complex Tasks

•       Despite scientific exigency

–          No precise and predictive nexuses for current theories of situation awareness and specific quantifiable cognitive and perceptual processes  

•       Theoretical Model of Situation Awareness (TMSA) recognizes perception as a critical first step, however, there are no specific or quantitative links between perceptual abilities and situation awareness

–          These are conceptual models that provide little specificity with regards to the neurocognitive processes

•       Needed theoretical advancement is hampered by lack of specific, testable predictions regarding plausible component processes

–          The results of this study will fill in a key gap in the TMSA, and in so doing produce an Enhanced – TMSAenabling both theoretical refinement and improved practical applications

•       Seminal research, has found that visual processing was involved in situation awareness. However, the specific neurocognitive component processes had not been explored in detail. This work must therefore test explicit hypotheses regarding specific abilities that may contribute to situation awareness.

•       Visual processing: can be categorized as static and dynamic, and may be operationalized as visual attentiveness and perceptiveness, integrated with visuospatial memory.  

•       Consequently, the primary purpose of this quasi-experimental quantitative study was to test the predictive value of these specific candidate visuo-cognitive abilities as predictors of situation awareness.

•       Therefore, this project tested the predictive value of these variables as factors between a particular task to be performed and the eventual outcome of situation awareness.

news summary (3)

Categories: Uncategorized

Weekly QuEST Discussion Topics and News, 7 Nov

November 6, 2014 Leave a comment

QuEST 7 Nov, 2014

The first topic this week is to revisit some of the material from the keynote given by Capt Amerika at the National Symposia for Sensor and Data Fusion.  Specifically last week Mike Y / Jared had questioned the use of some of the terms.  Capt A sent out a modified set of definitions and there was some email exchanges this week on the content.  So we want to take a couple of minutes updating the group and having any discussion on how we will use the terms agent, data, information, meaning, query, unexpected query, situation/simulation/structural coherence as defining characteristics of consciousness, consciousness and finally QuEST agents.

The second topic is once we’ve converged on those defining characteristics of consciousness we need to revisit specific driver problems to talk about engineering solutions.  Capt A met with Mike R this week and we spent the time to get into the details of his cyber effort.  We use those details to attempt to put some specificity on engineering artificially conscious solutions.  We hope to discuss ideas on ‘How to we generate a situated / simulated / structurally coherent representation for his cyber malicious activity detection driver problem’.

That discussion led us back to revisit the information in the ‘unreasonable effectiveness of data’ article.  There are some ideas in that article that provide us some insights into how to define the axes of the simulation in our artificial conscious agents using modern approaches to Big Data instead of waiting for large ‘truthed’ data sets.

news summary (8)

Categories: Uncategorized