Research Interests: David W. Aha
Research Interests
Our group's research focuses on the design, implementation, and
testing of intelligent agents for decision aids of interest to
the Navy/DoD. Our interests include case-based reasoning (CBR),
machine learning (e.g., transfer learning, active learning,
statistical relational learning, deep learning), meta-reasoning (e.g.,
goal reasoning), planning, mixed-initiative reasoning (e.g.,
conversational CBR), text analysis (e.g., of social media), and
related topics. Some of our current projects concern (2013 Feb):
- Autonomy and Meta-Reasoning: Goal reasoning (e.g., the goal-driven autonomy model)
allows agents to automatically respond to unexpected events in complex environments,
including self-selection of their goals
- Active Transfer Learning: Evaluation support and event discovery
for goal-driven autonomy
- Recursive Deep Learning: Automated scene interpretation
using constrained recursive deep learning techniques, where the
constraints ensure that situational awareness questions are addressed
- Model Interactions for Trusted Decision Systems: Learning the
models of users operating an unmanned vehicle simulator, and
gaining the trust of these operators
- Foundations of Web Analytics: Classifying users by their
browsing behavior
- Chat Attention Management: Automated recognition of urgent
chat messages and summarization of chat room content
(this pertains to our AAAI-11 Workshop on Analyzing Microtext)
For more information, please see the following links or
contact me.
Last updated: February 2013