Filters








1,820 Hits in 5.7 sec

Reasoning about Actions and State Changes by Injecting Commonsense Knowledge

Niket Tandon, Bhavana Dalvi, Joel Grus, Wen-tau Yih, Antoine Bosselut, Peter Clark
2018 Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing  
., a paragraph describing photosynthesis, requires modeling actions and the state changes they produce, so that questions about entities at different timepoints can be answered.  ...  In this paper, we show how the predicted effects of actions in the context of a paragraph can be improved in two ways: (1) by incorporating global, commonsense constraints (e.g., a non-existent entity  ...  Acknowledgements We thank Oren Etzioni for his insightful feedback and encouragement for this work. We are grateful to Paul Allen whose long-term vision continues to inspire our scientific endeavors.  ... 
doi:10.18653/v1/d18-1006 dblp:conf/emnlp/TandonDGYBC18 fatcat:fis5t7q2hbeznjq666v2tro3eu

Reasoning about Actions and State Changes by Injecting Commonsense Knowledge [article]

Niket Tandon, Bhavana Dalvi Mishra, Joel Grus, Wen-tau Yih, Antoine Bosselut, Peter Clark
2018 arXiv   pre-print
., a paragraph describing photosynthesis, requires modeling actions and the state changes they produce, so that questions about entities at different timepoints can be answered.  ...  In this paper, we show how the predicted effects of actions in the context of a paragraph can be improved in two ways: (1) by incorporating global, commonsense constraints (e.g., a non-existent entity  ...  Acknowledgements We thank Oren Etzioni for his insightful feedback and encouragement for this work. We are grateful to Paul Allen whose long-term vision continues to inspire our scientific endeavors.  ... 
arXiv:1808.10012v1 fatcat:xbfhtkcndvgmzmm4n34cqzaixi

Commonsense Knowledge Reasoning and Generation with Pre-trained Language Models: A Survey [article]

Prajjwal Bhargava, Vincent Ng
2022 arXiv   pre-print
This paper presents a survey of these tasks, discusses the strengths and weaknesses of state-of-the-art pre-trained models for commonsense reasoning and generation as revealed by these tasks, and reflects  ...  While commonsense knowledge acquisition and reasoning has traditionally been a core research topic in the knowledge representation and reasoning community, recent years have seen a surge of interest in  ...  Predicting object states after an action has been taken drastically improves a PLM's ability to make correct inferences about object states.  ... 
arXiv:2201.12438v1 fatcat:tpqwhagvnvdzlhzmbm6ajhokcm

EventNet: Inferring Temporal Relations Between Commonsense Events [chapter]

Jose Espinosa, Henry Lieberman
2005 Lecture Notes in Computer Science  
It comprises 10,000 nodes and 30,000 temporal links mined from the Openmind Commonsense Knowledge Base.  ...  The temporal relation might be one of cause and effect, of action/goal or prerequisite relations, or simply that they tend to follow each other in a commonly occurring "script".  ...  We believe we have shown that merely by generating a set of plausible possibilities for event occurrence, and using Commonsense reasoning, application context and user interaction to choose from those  ... 
doi:10.1007/11579427_7 fatcat:gkp2vh5nsrhhfpvgto4jo4xkte

Knowledge-Aware Procedural Text Understanding with Multi-Stage Training [article]

Zhihan Zhang, Xiubo Geng, Tao Qin, Yunfang Wu, Daxin Jiang
2020 arXiv   pre-print
Two challenges, difficulty of commonsense reasoning and data insufficiency, still remain unsolved.  ...  Specifically, we retrieve informative knowledge triples from ConceptNet and perform knowledge-aware reasoning while tracking the entities.  ...  Besides, entities usually undergo implicit state changes and commonsense knowledge is often required in reasoning.  ... 
arXiv:2009.13199v1 fatcat:dbh3yip5mvdpvim2hjoohm6eoi

Explaining Arguments with Background Knowledge

Maria Becker, Ioana Hulpuş, Juri Opitz, Debjit Paul, Jonathan Kobbe, Heiner Stuckenschmidt, Anette Frank
2020 Datenbank-Spektrum  
We introduce the problems and challenges we need to address, and present the progress we achieved until now by applying advanced natural language and knowledge processing methods.  ...  This requires human background knowledge and reasoning capacity, in order to explicate the complete reasoning of an argument.  ...  not explicitly stated in arguments, (ii) by exploiting static background knowledge sources that characterize the identified types of implicit knowledge relations between entities and concepts and (iii)  ... 
doi:10.1007/s13222-020-00348-6 fatcat:zd55bxjr7bhs5ab5whi3ih4q4y

Understanding in Artificial Intelligence [article]

Stefan Maetschke and David Martinez Iraola and Pieter Barnard and Elaheh ShafieiBavani and Peter Zhong and Ying Xu and Antonio Jimeno Yepes
2021 arXiv   pre-print
To answer these questions, we have analysed existing benchmarks and their understanding capabilities, defined by a set of understanding capabilities, and current research streams.  ...  A question remains of how much understanding is leveraged by these methods and how appropriate are the current benchmarks to measure understanding capabilities.  ...  This is an important task because human-human communications often describe the expected change of state after an action, and the actions to achieve the change are not detailed.  ... 
arXiv:2101.06573v1 fatcat:nlp6h5toh5f6lpwjsafn6gulbq

ToolNet: Using Commonsense Generalization for Predicting Tool Use for Robot Plan Synthesis [article]

Rajas Bansal, Shreshth Tuli, Rohan Paul, Mausam
2021 arXiv   pre-print
Can a robot learn commonsense knowledge and adapt to novel settings where some known tools are missing, but alternative unseen tools are present?  ...  This dataset maintains user plans involving multi-step object interactions along with symbolic state changes.  ...  Mausam is supported by an IBM SUR award, grants by Google, Bloomberg and 1MG, Jai Gupta chair fellowship, and a Visvesvaraya faculty award by Govt. of India. We thank Puig et al.  ... 
arXiv:2006.05478v3 fatcat:fz5iyw5ujbefvikjm6qzeyrdou

A Heterogeneous Graph with Factual, Temporal and Logical Knowledge for Question Answering Over Dynamic Contexts [article]

Wanjun Zhong, Duyu Tang, Nan Duan, Ming Zhou, Jiahai Wang, Jian Yin
2020 arXiv   pre-print
In this work, we propose a graph-based approach, where a heterogeneous graph is automatically built with factual knowledge of the context, temporal knowledge of the past states, and logical knowledge that  ...  combines human-curated knowledge bases and rule bases.  ...  Specifically, a graph conveys three kinds of knowledge, including factual knowledge about the context which is obtained by information extraction, temporal knowledge about the past states of participants  ... 
arXiv:2004.12057v1 fatcat:anocggngtzdqzclgkna42w4muy

Tiered Reasoning for Intuitive Physics: Toward Verifiable Commonsense Language Understanding [article]

Shane Storks, Qiaozi Gao, Yichi Zhang, Joyce Chai
2022 arXiv   pre-print
The TRIP dataset and our baseline results will motivate verifiable evaluation of commonsense reasoning and facilitate future research toward developing better language understanding and reasoning models  ...  Toward this goal, we introduce Tiered Reasoning for Intuitive Physics (TRIP), a novel commonsense reasoning dataset with dense annotations that enable multi-tiered evaluation of machines' reasoning process  ...  Acknowledgements This work was supported in part by the National Science Foundation (IIS-1617682 and IIS-1949634) and the DARPA XAI program through UCLA (N66001-17-2-4029).  ... 
arXiv:2109.04947v3 fatcat:imfjkmzfxng4xo26nfjaw5h3gm

"I'm Not Mad": Commonsense Implications of Negation and Contradiction [article]

Liwei Jiang, Antoine Bosselut, Chandra Bhagavatula, Yejin Choi
2021 arXiv   pre-print
Natural language inference requires reasoning about contradictions, negations, and their commonsense implications.  ...  We introduce ANION1, a new commonsense knowledge graph with 624K if-then rules focusing on negated and contradictory events.  ...  This research was supported in part by DARPA under the MCS program through NIWC Pacific (N66001-19-2-4031), and the Allen Institute for AI (AI2).  ... 
arXiv:2104.06511v2 fatcat:yqjnu3tnknc5bhjv42vxkf6av4

Logic meets Probability: Towards Explainable AI Systems for Uncertain Worlds

Vaishak Belle
2017 Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence  
Formal languages allow their internal properties to be robustly scrutinized, can be augmented by adding new knowledge, and are amenable to abstractions, all of which are vital to the design of intelligent  ...  Logical AI is concerned with formal languages to represent and reason with qualitative specifications; statistical AI is concerned with learning quantitative specifications from data.  ...  Nonetheless, given queries about outcomes after actions, as needed for automated planning, we have shown that actions can be either "compiled" away, or "injected" to the current state [Belle and Levesque  ... 
doi:10.24963/ijcai.2017/733 dblp:conf/ijcai/Belle17 fatcat:xrnwnlyhdzbkdhc23g76ysnaky

Delphi: Towards Machine Ethics and Norms [article]

Liwei Jiang, Jena D. Hwang, Chandra Bhagavatula, Ronan Le Bras, Maxwell Forbes, Jon Borchardt, Jenny Liang, Oren Etzioni, Maarten Sap, Yejin Choi
2021 arXiv   pre-print
Our prototype model, Delphi, demonstrates strong promise of language-based commonsense moral reasoning, with up to 92.1% accuracy vetted by humans.  ...  natural language descriptions; (3) commonsense reasoning to anticipate the outcome of alternative actions in different contexts; (4) most importantly, the ability to make ethical judgments given the interplay  ...  This research was supported in part by DARPA under the MCS program through NIWC Pacific (N66001-19-2-4031), and the Allen Institute for AI (AI2).  ... 
arXiv:2110.07574v1 fatcat:yi4kplseqfanfccina227nx2zi

Teaching Knowledge: the lights that teachers live by

Margret Buchmann
1987 Oxford Review of Education  
determining what each kind of teaching knowledge tends to be about, how it arises and changes, how it is held and used, what it teaching, on the whole, account for the aspects by which teaching is viewed  ...  How bright or dim are these different lights, and how are old lights in teaching changed by new ones, if they change at all?  ... 
doi:10.1080/0305498870130203 fatcat:66pdk6sg5zfithaxji44a4ucre

Update by means of inference rules [chapter]

Teodor C. Przymusinski, Hudson Turner
1995 Lecture Notes in Computer Science  
Katsuno and Mendelzon have distinguished two abstract frameworks for reasoning about change: theory revision and theory update.  ...  Theory revision involves a change in knowledge or belief with respect to a static world. By contrast, theory update involves a change of knowledge or belief in a changing world.  ...  We are also grateful for comments from Chitta Baral, Enrico Giunchiglia and G. N. Kartha. The rst author is partially supported by NSF grant #IRI-9313061; the second by NSF grant #IRI-9306751.  ... 
doi:10.1007/3-540-59487-6_12 fatcat:bo5z6ne3xnezzofse5fslrtgky
« Previous Showing results 1 — 15 out of 1,820 results