Making Sense of Sensed Data - Issues in Developing Heterogeneous Cognitive Systems (MSSD15)

In conjunction with Computer Frontiers 2015

A new frontier in computing is evolving around developing systems that 'sense' and respond to data in ways that represent or mimic human 'cognition'. These advancements represent the intersection of developments in processor hardware, software, and sensing devices. Examples include the implementation of low power architectures that mimic the type of synaptic information transmission that occur in animal brains (Neuromorphic systems), improvements in probability based software for big data analysis systems such as Natural Language Processing, the development of "deep learning systems" which have demonstrated systems that 'learn' representations of data, and the explosive growth of user based content in data systems such as cell phone video etc.. The evolution of these systems is leading to new possibilities for computing systems to interpret a vast amount of big data collected by sensors into a meaningful representations and to learn from that data or draw inferences from it.

This workshop is intended to spur discussion and highlight the possibilities, opportunities, and challenges in the evolution of these new systems that begin to perceive the world,

Focus areas of the workshop will include:

Hardware
  1. Developments in Neuromorphic logic architectures, what are the benefits and drawbacks?
  2. Are logic models necessary to be based on biological neural models, or can Von-neuman architectures satisfy the requirement for cognitive systems?
  3. Advancements sensor implementations that represent data and their growth in the "internet of things" (IoT)
Software
  1. Are there a set of software tools or platforms that are early leaders in the area?
  2. How will sensing cognitive systems be evaluated? Against human perception? What is a learning system?
  3. Traditional AI machine learning vs. deep neural representations
Ecosystem
  1. Evolving business cases for cognitive architecture systems
  2. Collaboration between governments, corporations, and universities. Who should lead?
  3. What industry shifts may occur as the result of systems better able to interpret sensed data?

Organization

Workshop Organizers: Committee Chair Lead Committee members

Key Dates

Submit Abstract by Friday, Feb 13, 2015 to grecon@us.ibm.com.

Notification of Acceptance : Wednesday Feb 23, 2015
Final Publication no more than 8 pages: March 10, 2015

At least one author per accepted paper, must be registered for the conference. There is one flat fee for both the conference and workshop.