Personal tools
You are here: Home Members IanWilson's Home

Ian Wilson

Meta Research into the functional requirements for reactive and emotional behavior and how they are affected by personality gender and age. How these behaviors affect facial gestures, body gestures and action selection.

Ian Wilson
Ian Wilson
Emotion AI
Key research interests:

Computational Neuroscience of emotional level systems and their functional representations. Personality, gender, age and functional variability in brain systems. Arousal, stress, anxiety. Facial and body gestures, motivation and action determination.

Definitions


"The emotion level brain systems evolved to prioritize events in the environment so a creature with limited energy and processing power could attend to those events that were the most important to it at that moment in time. The system is adaptive in that it can optimize its parameters of operation to better match its environment in real time."

"Personality is the variation in configuration and performance of the reactive, emotional and cognitive brain systems. It confers an advantage on a species by widening the possible behavior space of that species by allowing a single brain system to be effectively parameterized so the single system can work in many different configurations. Individuals within a species can find niches within their group and thus personality is also adaptive within the group."


Open Questions


In modeling the systems I am researching I have been able to make many interesting observations and it has presented many questions for which I currently see no definative answers.  My current primary research questions are these:

  1. What are the primary functions involved with emotion level processing
  2. What are the primary brain structures
  3. What are the primary hormonal messengers
  4. What are the primary neurotransmitters both exitory and inhibitory
  5. How do arousal, anxiety and stress fit into the model and does the exitation and inhibition of these factors define primary emotional functioning?

My current model splits the traditional 3 dimensional personality/emotion model into an essentially n dimensional "signature" (behavioral functions with both exitory and inhibitory elements arranged on a 2 dimensional axis - value X element). I call this a "behavior signature" or in my flights of fantasy a "wilson behavior signature".

I also define this model within a 4 level hierachy which represents - reactions (stimulus / response learning),  emotional behavior (associative learning), cognition (symbolic learning) and conciousness (semantic learning). Here each higher level monitors the previous and each lower level "informs" the higher levels of its state. 



Structural Problems in Research


During the course of my research I have encountered many structural problems with research in general that I am sure many of you also encounter day in and day out. This web portal is a great way of addressing some of those issues but by way of hoping to assist science in general I wanted to list some of the issues I have found and hear from others if these really are problems and if so how can they be fixed:

  • The increasing commercialization of research. I am not reffering to commercial research (from which I pay my rent) but to public research for which the public (you and I) are rapidly being denied access as it is being locked up by event organizers and publishers. This to my mind is unacceptable. A web search 5 years ago would result in a wealth of research papers being returned but now all one can find is abstracts, the actual papers having been copyrighted by Big Publishing Corp and hidden from the non paying public. In a field like emotion where we cross many, many scientific fields this is a very large burden. If I were to subscribe to all of the journals and associations required to get access to the research papers I need to do my work I estimate it would cost me upwards of $50,000, per year.
  • Lack of  central structure. As researchers we can waste a great deal of time "reinventing the wheel" by researching and trying to understand areas that other researchers have understood, written about and moved on from, simply because we are unaware of their work. Especially in a cross disciplinary field like this one it is very difficult to know the state of the art in any sub field. However even within researchers own fields I see continual wheel reinvention. The problem of course is that there is no central or authoritative "tree of knowledge" to which we can refer, but hopefully efforts like this web site can make a start. I am sure everyone looking at my list of open questions will see something they know has already been comprehensively covered by others.
  • Lack of cross pollenation. I recently attended a cross disciplinary conference and noted that many of the paper references were generally from the single field that the researcher was in, so AI people might reference exclusively other AI researchers when talking about architectures for neuroscience but not reference neuroscientists. Naturally there is only so much time in the day to work but again, linked to the above point, this introduces alot of reinvention of the wheel. It would benefit all of the different connected fields if we had some way of working together to validate and strengthen each others work (which was one of the points I believe of the conference I attended).

I would very much like to see the democtratization, dissemination and structuring of research (which I fully admit is hypocritical for me as my research is commercial and I struggle with the social consequences of this). To that end, and similar in a way to this portal, is my solution:

  • Reasearch Blogs.  Here researchers blog their research on a day by day basis so others can see and comment on it, suggesting improvements or areas that others have already covered. Creating, like this special area we have, a community of researchers in any particular field. Aggregation sites could collect all current ongoing research in one place for easy reading and appraisal.
  • Research Paper Aggregation. When the researcher is ready to put their current ideas together in a paper those papers would be posted to a field/paper aggregation site where "the whole field" can judge its merits. These paper aggregation sites would be hierarchical going from at the lowest level very specific sub fields through to something like "Human Behavior" for this area and then onto the top level "Nature/Science" type site. Papers can be progressively aggregated up the chain where judgement is by the whole community. This community approach not only frees the research but allows the community to help each other determine the state of the art in any field, saving us all time and energy.

Any comments and or help are welcome.





Document Actions
Powered by Plone

Portal usage statistics