Probabilistic Action Cores
| Version | 1.1.2 |
|---|---|
| Framework | Python |
| Type | Natural-language instruction interpreter |
| License | BSD |
| Lead Developer | Daniel Nyga |
| Institute | Institute for Artificial Intelligence, University of Bremen |
| Website | http://www.actioncores.org |
PRAC (Probabilistic Action Cores) is an
German Research Foundation (DFG).[1]
Goals
The ultimate goal of the PRAC system is to make knowledge about everyday activities from websites like
probabilistic relational models, PRAC uses the principles of analogical reasoning and instance-based learning to infer completions of roles in semantic networks.[4]
PRAC has been successfully applied to teach robots to conduct chemical experiments[5] and to make pancakes and pizza from wikiHow articles.[6]
References
- ^ Nyga, Daniel (2017). "Interpretation of Natural-language Robot Instructions: Probabilistic Knowledge Representation, Learning, and Reasoning" (PDF). PhD Thesis.
- S2CID 302048.
- ^ Nyga, Daniel; Beetz, Michael (2015). "Cloud-based Probabilistic Knowledge Services for Instruction Interpretation" (PDF). International Symposium of Robotics Research (ISRR).
- ^ Nyga, Daniel; Picklum, Mareike; Koralewski, Sebastian; Beetz, Michael (2017). "Instruction Completion through Instance-based Learning and Semantic Analogical Reasoning". International Conference on Robotics and Automation (ICRA).
- S2CID 7613082.
- ^ Will Knight (August 24, 2015). "Robots Learn to Make Pancakes from WikiHow Articles". MIT Tech Review. Retrieved 2017-03-14.
A robot called PR2 in Germany is learning to prepare pancakes and pizzas by carefully reading through WikiHow's written directions.