scispace - formally typeset
Search or ask a question
Topic

Abductive reasoning

About: Abductive reasoning is a research topic. Over the lifetime, 1917 publications have been published within this topic receiving 44645 citations. The topic is also known as: abduction & abductive inference.


Papers
More filters
Proceedings ArticleDOI
01 Apr 2019
TL;DR: The presented approach is based on the OODA loop and implemented within the framework of the Transferable Belief Model, Choquet Expected Utility, Multiple Quantile Model, abductive reasoning utilizing the Belief-based argumentation, and the Anytime decision making.
Abstract: This paper addresses the problem of situation management in an uncertain and ambiguous environment. It presents an approach to reasoning and decision making for identifying asymmetric threats and preventing or mitigating possible catastrophic consequences under resource constraints. Decision making for action selection incorporates decision makers’ attitude toward risk and uncertainty in the situations involving regular as well as high consequence rare events. The presented approach is based on the OODA loop and implemented within the framework of the Transferable Belief Model, Choquet Expected Utility, Multiple Quantile Model, abductive reasoning utilizing the Belief-based argumentation, and the Anytime decision making. A quality control component is added to the OODA loop to deal with possible insufficient quality of input information and results of processes involved. An example illustrating the approach is presented in this paper.

6 citations

01 May 1990
TL;DR: A probabilistic model of text understanding is developed, devised a way of incrementally constructing and evaluating belief networks that is applicable to other abduction problems and treated all aspects of the text-understanding problem in a unified way.
Abstract: We discuss a new framework for text understanding. Three major design decisions characterize this approach. First, we take the problem of text understanding to be a particular case of the general problem of abductive inference: reasoning from effects to causes. Second, we use probability theory to handle the uncertainty that arises in abductive inference in general and natural language understanding in particular. Finally, we treat all aspects of the text-understanding problem in a unified way. All aspects of natural language processing are treated in the same framework, allowing us to integrate syntactic, semantic and pragmatic constraints. In order to apply probability theory to this problem, we have developed a probabilistic model of text understanding. To make it practical to use this model, we have devised a way of incrementally constructing and evaluating belief networks that is applicable to other abduction problems. We have written a program, Wimp3, to experiment with this framework.

6 citations

Journal Article
TL;DR: In this paper, the authors explored the cases and features of the abductive inference used by young children when trying to explain natural phenomena in everyday life and found that even young children use abductive inferences to explain doubtful natural phenomena.
Abstract: The purpose of this study is to explore the cases and features of the abductive inference used by young children when trying to explain natural phenomena in everyday life. From observing a 5-year-old's daily activities with his family, and analyzing the data according to the criterion extracted from the form of abductive inference described by C. S. Peirce, a few cases where the child used abductive inferences to explain natural phenomena were found. The abductive inferences in the cases were conducted: (a) based on figural resemblance and behavioral resemblance (b) under the influence by individual belief and communal belief, then (c) resulted in new categorization accompanied by over generalization. Such features of the abductive inference showed the 'double faces'; sometimes encourages and sometimes discourages children's generating better scientific hypotheses and explanations. These results suggest that even young children use abductive inference to explain doubtful natural phenomena in everyday life, although we need to consider carefully with the double aspects of the features of abductive inference for the practical applications to the fields of science education. Finally, several suggestions and following studies for science education are proposed.

6 citations

Journal ArticleDOI
30 Jun 2015
TL;DR: In this paper, a variety of analyses have been performed on the different modes of scientific thinking; they mainly revise the use of what has been called logical rationality and narrative rationality in science education in the different educational levels, from Kindergarten to University, under the hypothesis that these two modes of thinking can be recognised in the diversity of scientific texts that are used when teaching science.
Abstract: In this paper I go through a variety of analyses that have been performed on the different ‘modes’ of scientific thinking; I do this from the perspective of didactics of science (i.e., science education as an academic discipline). I mainly revise the use of what has been called ‘logical rationality’ and ‘narrative rationality’ in science education in the different educational levels, from Kindergarten to University, under the hypothesis that these two modes of thinking can be recognised in the diversity of scientific texts that are used when teaching science. I relate the first mode of thinking to the ‘syntactic’ structure of the scientific disciplines, validated within the famous ‘context of justification’, and the second mode to the historical development of the disciplines, which occurs in the ‘context of discovery’. I recognise, in science classes and textbooks, a third, ‘hybrid’ mode of thinking, which can prove fruitful under the light of the current imperatives of science education for all. I propose that this new mode is based on the use of abductive reasoning, which ‘generates’ hypotheses and strongly employs theoretical models. Focussing on the ways in which models and evidence are used, I analogue scientific thinking to detective thinking; in both, evidence plays a central role when projecting the model to the problem to be solved.

6 citations

Proceedings ArticleDOI
25 Jul 2020
TL;DR: In this paper, a learning-to-rank framework was proposed to evaluate the abductive reasoning ability of a learning system, where two observations are given and the most plausible hypothesis is asked to pick out from the candidates.
Abstract: The abductive natural language inference task (αNLI) is proposed to evaluate the abductive reasoning ability of a learning system. In the αNLI task, two observations are given and the most plausible hypothesis is asked to pick out from the candidates. Existing methods simply formulate it as a classification problem, thus a cross-entropy log-loss objective is used during training. However, discriminating true from false does not measure the plausibility of a hypothesis, for all the hypotheses have a chance to happen, only the probabilities are different. To fill this gap, we switch to a ranking perspective that sorts the hypotheses in order of their plausibilities. With this new perspective, a novel L2R2 approach is proposed under the learning-to-rank framework. Firstly, training samples are reorganized into a ranking form, where two observations and their hypotheses are treated as the query and a set of candidate documents respectively. Then, an ESIM model or pre-trained language model, e.g. BERT or RoBERTa, is obtained as the scoring function. Finally, the loss functions for the ranking task can be either pair-wise or list-wise for training. The experimental results on the ART dataset reach the state-of-the-art in the public leaderboard.

6 citations


Network Information
Related Topics (5)
Natural language
31.1K papers, 806.8K citations
82% related
Ontology (information science)
57K papers, 869.1K citations
79% related
Inference
36.8K papers, 1.3M citations
76% related
Heuristics
32.1K papers, 956.5K citations
76% related
Social network
42.9K papers, 1.5M citations
75% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
202356
2022103
202156
202059
201956
201867