Communicating with Computers

People Involved:

  • Mayukh Das
  • Shuo Yang
  • Phillip Odom

The grand vision of smart machines is enabling AI agents and humans to interact seamlessly, make decisions and solve problems together and learn from as well as complement each others capabilities. We share such a vision and we build frameworks and techniques to leverage human knowledge in creating smarter sequential decision-making and predictive systems and protocols for such knowledge elicitation.

Our research includes, but is not limited to, knowledge-augmented Statistical Relational Learning, human guided and collaborative decision-making and planning (esp. in stochastic, partially observable, semi-structured environments), “active” human-AI interaction, various modalities of human guidance and, finally, successful application of such systems to real-world tasks such as Health, Biomedicine and Finance.

Concept learning, domain transfer/extension and higher level knowledge induction are some additional interesting research ventures (motivated by DARPA ‘Communicating with Computers’). We aim for an over-arching framework for human-AI collaboration involving multiple human (non)experts and machines with varied modalities of interaction.

References:

  • Das, M., Odom, P., Islam, M.R., Doppa, J., Roth, D., & Natarajan, S., “Preference- Guided Planning: An Active Elicitation Approach”, International Conference on Autonomous Agents and Multiagent Systems (AAMAS) 2018.
  • Das, M., Islam, M.R., Doppa, J.R., Roth, D., & Natarajan, S., “Active Preference Elicitation for Planning”, Human-Machine Collaborative Learning workshop(@ AAAI) 2017.
  • Narayan-Chen A., Graber C., Das M., Islam M.R., Dan S., Natarajan S., Doppa J.R., Hockenmaier J., Palmer M., Roth D., “Towards Problem Solving Agents that Communicate and Learn.” Workshop on Language Grounding for Robotics at ACL 2017.
  • Alexander L. Hayes, Mayukh Das, Phillip Odom, Sriraam Natarajan. “User Friendly Automatic Construction of Background Knowledge: Mode Construction from ER Diagrams.” Knowledge Capture Conference 2017.
  • Odom, P., & Natarajan, S., “Active Advice Seeking for Inverse Reinforcement Learning”, International Conference on Autonomous Agents and Multiagent Systems (AAMAS) 2016.
  • Odom, P., & Natarajan, S., “Actively Interacting with Experts: A Probabilistic Logic Approach”, European Conference on Machine Learning and Principles of Knowledge Discovery in Databases (ECMLPKDD) 2016.
  • Odom, P., Kumaraswamy, R., Kersting, K., & Natarajan, S., “Learning through Advice-Seeking via Transfer”, International Conference on Inductive Logic Programming (ILP) 2016.
  • Odom, P., Khot, T., Porter, R., & Natarajan, S., “Knowledge-Based Probabilistic Logic Learning”, Twenty-Ninth AAAI Conference on Artificial Intelligence (AAAI) 2015.
  • Odom, P., Bangera, V., Khot, T., Page, D., & Natarajan, S., “Extracting Adverse Drug Events from Text using Human Advice”, Artificial Intelligence in Medicine (AIME) 2015.
  • Yang, S., Khot, T., Kersting, K., Kunapuli, G., Hauser, K., & Natarajan, S., “Learning from Imbalanced Data in Relational Domains: A Soft Margin Approach”, International Conference on Data Mining (ICDM) 2014.
  • Yang, S., & Natarajan, S., “Knowledge Intensive Learning: Combining Qualitative Constraints with Causal Independence for Parameter Learning in Probabilistic Models”, European Conference on Machine Learning, (ECMLPKDD) 2013.

Updated: