2,445 Hits in 2.7 sec

The Semantic Understanding of the English Counterfactual Conditionals—A Model Based on Conceptual Integration Theory

Ling Qin
2013 Journal of Language Teaching and Research  
better model of explaining the semantic understanding of the English Counterfactual Conditionals.  ...  Anyway, from a cognitive perspective, Fauconnier (1997) language is a superficial manifestation of hidden, highly abstract, and cognitive constructions (p.34) English Counterfactual Conditionals 1 , as  ...  CI has powerful explanatory capability in reasoning including counterfactual reasoning.  ... 
doi:10.4304/jltr.4.4.754-766 fatcat:a5reqsjzljf6rhe5yqvsbxpml4

CLEVRER: CoLlision Events for Video REpresentation and Reasoning [article]

Kexin Yi, Chuang Gan, Yunzhu Li, Pushmeet Kohli, Jiajun Wu, Antonio Torralba, Joshua B. Tenenbaum
2020 arXiv   pre-print
reasoning should incorporate the capability of both perceiving complex visual and language inputs, and understanding the underlying dynamics and causal relations.  ...  Most video reasoning benchmarks, however, focus on pattern recognition from complex visual and language input, instead of on causal structure.  ...  Video understanding. With the availability of large-scale video datasets (Caba Heilbron et al., 2015; Kay et al., 2017), joint video and language understanding tasks have received much interest.  ... 
arXiv:1910.01442v2 fatcat:5t3rceq24rffjkdi4nmongam3q

Semantic Modeling for Food Recommendation Explanations [article]

Ishita Padhiar, Oshani Seneviratne, Shruthi Chari, Daniel Gruen, Deborah L. McGuinness
2021 arXiv   pre-print
Our motivation with the use of FEO is to empower users to make decisions about their health, fully equipped with an understanding of the AI recommender systems as they relate to user questions, by providing  ...  Such explanations would benefit users of recommendation systems by empowering them with justifications for following the system's suggestions.  ...  Then the system generates the persuasion message with explanations using natural language templates [20] .  ... 
arXiv:2105.01269v1 fatcat:wmpfudjcl5gprk6max2sj7wi6u

Explainable AI and Multi-Modal Causability in Medicine

Andreas Holzinger
2021 I-COM: A Journal of Interactive and Cooperative Media  
The key for future human-AI interfaces is to map explainability with causability and to allow a domain expert to ask questions to understand why an AI came up with a result, and also to ask "what-if" questions  ...  (counterfactuals) to gain insight into the underlying independent explanatory factors of a result.  ...  Explicit Knowledge can be explained, e. g. by articulating it via natural language etc. and can be shared with other people.  ... 
doi:10.1515/icom-2020-0024 fatcat:bfmqrfmgvzatxpds73lkuri3nm

Counterfactual Vision-and-Language Navigation via Adversarial Path Sampling [article]

Tsu-Jui Fu, Xin Eric Wang, Matthew Peterson, Scott Grafton, Miguel Eckstein, William Yang Wang
2020 arXiv   pre-print
We propose an adversarial-driven counterfactual reasoning model that can consider effective conditions instead of low-quality augmented data.  ...  Vision-and-Language Navigation (VLN) is a task where agents must decide how to move through a 3D environment to reach a goal by grounding natural language instructions to the visual surroundings.  ...  Introduction Vision-and-language navigation (VLN) [3, 8] is a complex task that requires an agent to understand natural language, encode visual information from the surrounding environment, and associate  ... 
arXiv:1911.07308v3 fatcat:dxyvp2ocmzalhegpahqlj4wi4i

Transparency as design publicity: explaining and justifying inscrutable algorithms

Michele Loi, Andrea Ferrario, Eleonora Viganò
2020 Ethics and Information Technology  
We criticize recent attempts to identify the explanation of black box algorithms with making their decisions (post-hoc) interpretable, focusing our discussion on counterfactual explanations.  ...  We argue that design publicity can be more easily linked with the justification of the use and of the design of the algorithm, and of each individual decision following from it.  ...  For this reason, making such translation a publicly verifiable criteria provides the public and scientific community with the information to assess how a given goal is operationalized in machine-language  ... 
doi:10.1007/s10676-020-09564-w pmid:34867077 pmcid:PMC8626372 fatcat:ryvyugfghzdylcwgop7dqbexmu

Meaningful Explanations of Black Box AI Decision Systems

Dino Pedreschi, Fosca Giannotti, Riccardo Guidotti, Anna Monreale, Salvatore Ruggieri, Franco Turini
Black box AI systems for automated decision making, often based on machine learning over (big) data, map a user's features into a class or a score without exposing the reasons why.  ...  the urgent open challenge of how to construct meaningful explanations of opaque AI/ML systems, introducing the local-toglobal framework for black box explanation, articulated along three lines: (i) the language  ...  Stakeholders should be empowered to reason on explanations, to understand how the automated decisionmaking system works on the basis of the inputs provided by the user; what are the most critical features  ... 
doi:10.1609/aaai.v33i01.33019780 fatcat:g5btddfrcvcuxerswzthvsliey

Explainability Pitfalls: Beyond Dark Patterns in Explainable AI [article]

Upol Ehsan, Mark O. Riedl
2021 arXiv   pre-print
To make Explainable AI (XAI) systems trustworthy, understanding harmful effects is just as important as producing well-designed explanations.  ...  Acknowledgments With our deepest gratitude, we thank our participants for generously investing their time in the case study.  ...  AI decisions with "zero friction" or understanding.  ... 
arXiv:2109.12480v1 fatcat:cm2aa23yifa7bnx4kh7omxgnsq

Causality Learning: A New Perspective for Interpretable Machine Learning [article]

Guandong Xu, Tri Dung Duong, Qian Li, Shaowu Liu, Xianzhi Wang
2021 arXiv   pre-print
This paper provides an overview of causal analysis with the fundamental background and key concepts, and then summarizes most recent causal approaches for interpretable machine learning.  ...  To empower the capability of counterfactual explanations, constraints are considered in optimization problem of counterfactual explanation.  ...  To better understand causal inference, we give the following example combined with the notations defined above.  ... 
arXiv:2006.16789v2 fatcat:ole3dvpnjnfkflldd6to4nrrwq

ESTER: A Machine Reading Comprehension Dataset for Event Semantic Relation Reasoning [article]

Rujun Han, I-Hung Hsu, Jiao Sun, Julia Baylon, Qiang Ning, Dan Roth, Nanyun Peng
2021 arXiv   pre-print
The dataset leverages natural language queries to reason about the five most common event semantic relations, provides more than 6K questions and captures 10.1K event relation pairs.  ...  reasoning.  ...  Therefore, for machines to achieve human-level narrative understanding, we need to test and ensure models' capability to reason over these event relations.  ... 
arXiv:2104.08350v2 fatcat:tzdhlcezcvbw5ggr2mtaxo3gky

Multi-Domain Transformer-Based Counterfactual Augmentation for Earnings Call Analysis [article]

Zixuan Yuan, Yada Zhu, Wei Zhang, Ziming Huang, Guangnan Ye, Hui Xiong
2021 arXiv   pre-print
Then, a multi-domain counterfactual learning framework is developed to evaluate the gradient-based variations after we perturb limited EC informative texts with plentiful cross-domain documents, enabling  ...  Meanwhile, these black-box methods possess inherent difficulties in providing human-understandable explanations.  ...  Recent studies develop the influence function (Koh and Liang 2017) or its hessian-free variant (Zhang et al. 2021) to understand the counterfactual effect of training points on a prediction.  ... 
arXiv:2112.00963v2 fatcat:gaqw6g4v3rab7m7vxe7c6j5kce

Sufficient Causes: On Oxygen, Matches, and Fires

Judea Pearl
2019 Journal of Causal Inference  
We contrast this demonstration with the potential outcome framework and address the distinction between causes and enablers.  ...  We demonstrate how counterfactuals can be used to compute the probability that one event was/is a sufficient cause of another, and how counterfactuals emerge organically from basic scientific knowledge  ...  inputs for causal analysis. (2) To empower researchers with methods of estimating counterfactuals directly from functional description of their problems.  ... 
doi:10.1515/jci-2019-0026 fatcat:cplzpeuvyndyvkpikciwa4e2e4

The Offense-Defense Balance of Scientific Knowledge: Does Publishing AI Research Reduce Misuse? [article]

Toby Shevlane, Allan Dafoe
2020 arXiv   pre-print
The language of "let the users decide for themselves", reminiscent of computer security discourse, would lose its empowering sentiment if users become landed with problems for which no good solution exists  ...  This contrasts with the case where an effective treatment can be developed within a reasonable time period, which could weigh in favour of publication [15] .  ... 
arXiv:2001.00463v2 fatcat:pxc4altffrgdzevh6244f5lmme

Facing Immersive "Post-Truth" in AIVR?

Nadisha-Marie Aliman, Leon Kester
2020 Philosophies  
Building on this, Section 4 briefly discusses how future affective computing and virtual reality methods could be harnessed for counterfactual and other measures that seek to allow an understanding and  ...  Thereby, human perception imposes cognitive-affective concepts on the world, often previously constructed in social reality (abbreviated with SR in the following) and shared via language.  ... 
doi:10.3390/philosophies5040045 fatcat:3adrjkgrdrbthpot7jsexofsle

Experimental jurisprudence

Roseanna Sommers
2021 Science  
Skeptics in this camp contend that proximate cause represents nothing more than judicial anarchy: Judges engage in outcome-driven reasoning, which they dress up, post hoc, in the language of causality  ...  For example, using survey experiments, psychologists have discovered that laypeople's causal judgments are affected by counterfactual reasoning, which is in turn influenced by whether an agent behaved  ... 
doi:10.1126/science.abf0711 pmid:34437107 fatcat:m7khzltnvfhe3iztlban7oecqe
« Previous Showing results 1 — 15 out of 2,445 results