|Home -> Research -> Knowledge Acquisition
|Knowledge acquisition concerns how to obtain knowledge from external sources. Motivated from the requirements of a domestic robot, we consider chance discovery and inductive learning problems.|
| Recent Projects|
|Induction in Nonmonotonic Causal Theories for Reasoning about Action|
| Nonmonotonic causal theories are devoted to be a nonmonotonic formalism for representing causal knowledge, which can be used to formalize action domains, including indirect effects of actions, implied action preconditions, concurrent actions, nondeterministic actions, ramification and qualification constraints. However, in practice, it is hard to model a complicated action domain completely and there always exist unexpected scenarios which are not coved by a specific theory. An important problem for developing an intelligent agent is how to automatically modify a causal theory for an action domain so that the updated theory can cover a new scenario once it is encountered during the agent's exploration of the real world.
To address the problem, we tackle the inductive learning problem in nonmonotonic causal theories in the setting of domestic service robots. We reduce the learning task into the problem of modifying a causal theory such that the interpretations corresponding to new scenarios become a model of the updated theory, while all the original models keep unchanged. Learned causal rules are generalized due to corresponding "causally relevance" relations. We illustrate the approach case studies based on "KeJia".
Xiaoping Chen, Jianmin Ji, Jiongkun Xie, and Zhiqiang Sui. Toward open knowledge enabling for human-robot interaction. Journal of Human-Robot Interaction 1(2):100–117. 2012.. .
Xiaoping Chen, Jianmin Ji, Zhiqiang Sui, and Jiongkun Xie. Handling Open Knowledge for Service Robots, Proceedings of IJCAI 13, Beijing, China, Aug 3-9, 2013. .
Xiaoping Chen. Handling Open Knowledge for Service Robots, invited talk at Tenth International Workshop on Nonmonotonic Reasoning, Action and Change (NRAC 2013), Beijing, China, Aug 5, 2013.. .
Xiaoping Chen. Enhancing Human-Robot Interaction with Open Knowledge, invited talk at the 2013 RoSEC International Summer Workshop on Recent advances in cognitive robotics and HRI, Seoul, South Korea Aug. 28-29, 2013.. .
Jianmin Ji and Xiaoping Chen. Induction in Nonmonotonic Causal Theories for a Domestic Service Robot, In: Proceedings of the 21st International Conference on Inductive Logic Programming (ILP 2011), United Kingdom, 31st July - 3rd August, 2011. .
Huan Li, Xin Li, Dawei Hu, Tianyong Hao, Liu Wenyin, Xiaoping Chen. Adaptation rule learning for case-based reasoning. Concurrency and Computation: Practice and Experience, Volume 21, Number 5, 673-689. 10 April 2009. .
Yang Yang and Xiaoping Chen. An approach to plan-recognition based on sequential events, Computer Engineering, 31(12), 2005. .
Xiaoping Chen. Cognitive Relevance and Chance Discovery. Apperared in Akinori Abe and Yukio Ohsawa (Eds), Readings in Chance Discovery, Advanced Knowledge International 2005. .
Shizhuo Zhu and Xiaoping Chen. A survey on chance discovery. Computer Science, Vol.31, No.2.1-5, Feb, 2004. .
Liang Pi and Xiaoping Chen. A Quantitative Method for Representing Impacts of Chance based on simplified MDP/POMDP models. The First European Workshop on Chance Discovery. Aug. 2004. .
Shizhuo Zhu and Xiaoping Chen. A relative description for chance discovery of agents. Computer Engineering and Applications, Vol.40, No.5. May, 2004. .
Shizhuo Zhu, Xiaoping Chen and Liang Pi. A characterization of agent chance discovery: Abduction and its extension. Computer Engineering, Vol.30, No.12, Dec, 2004. .
Shizhuo Zhu and Xiaoping Chen. Integrating Abductive Reasoning with Lm4c for Chance specification and Chance Discovery Inference. Proc of HCI- 2003 Workshop on CD&CM. Greece, Jun 2003. .
Shizhuo Zhu and Xiaoping Chen. Relevance in Chance Discovery. Proc of HCI- 2003 Workshop on CD&CM. Greece, Jun 2003.. .
Liang Pi and Xiaoping Chen, et al. A machinery of chance discovery based on KDD. Proceedings of 20th National Conference on Databases. 2003. .
Liang Pi and Xiaoping Chen. Twofold-relative schema for chance discovery. Proceedings of 10th Notional Conference on Artificail Intelligence. Nov, 2003. .
C. Ji and X. Chen. Chance Discovery in a BDI Perspective of Planning: Anticipation, Participation, and Correlation. In: Proceedings of the 2nd International Workshop on Chance Discovery. Oct. 2002. Tokyo. .