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
Journal ArticleDOI
TL;DR: In this paper, the authors argue that machine learning techniques can be very useful in theory construction during a key step of inductive theorizing, which is called algorithm supported induction, yielding conclusions about patterns in data that are likely to be robustly replicated by other analysts and in other samples from the same population.
Abstract: Across many fields of social science, machine learning (ML) algorithms are rapidly advancing research as tools to support traditional hypothesis testing research (e.g. through data reduction and automation of data coding, or for improving matching on observable features of a phenomenon or constructing instrumental variables). In this Organization Science Perspective-paper, we argue that researchers are yet to recognize the value of ML techniques for theory building from data. This may be in part due to scholars’ inherent distaste for “predictions without explanations” that ML algorithms are known to produce. However, precisely because of this property, we argue that ML techniques can be very useful in theory construction during a key step of inductive theorizing—pattern detection. ML can facilitate “algorithm supported induction,” yielding conclusions about patterns in data that are likely to be robustly replicable by other analysts and in other samples from the same population. These patterns can then be used as inputs to abductive reasoning for building or developing theories that explain them. We propose that algorithm supported induction is valuable for researchers interested in using quantitative data to both develop and test theories in a transparent and reproducible manner, and we illustrate our arguments using simulations.

6 citations

Journal ArticleDOI
TL;DR: In this article, a query-driven, top-down execution model for predicate answer set programming with constraints is proposed to model commonsense reasoning with a sound, logical basis using Event Calculus (EC).
Abstract: Automated commonsense reasoning is essential for building human-like AI systems featuring, for example, explainable AI. Event Calculus (EC) is a family of formalisms that model commonsense reasoning with a sound, logical basis. Previous attempts to mechanize reasoning using EC faced difficulties in the treatment of the continuous change in dense domains (e.g., time and other physical quantities), constraints among variables, default negation, and the uniform application of different inference methods, among others. We propose the use of s(CASP), a query-driven, top-down execution model for Predicate Answer Set Programming with Constraints, to model and reason using EC. We show how EC scenarios can be naturally and directly encoded in s(CASP) and how it enables deductive and abductive reasoning tasks in domains featuring constraints involving both dense time and dense fluents.

6 citations

Proceedings ArticleDOI
18 Jun 2010
TL;DR: A theory of narrative functions that serve as a heuristic for relevance in narrative, and evidence that this heuristic is effective for disambiguation that leads to consistent understanding is provided.
Abstract: Story understanding requires a degree of knowledge and expressiveness beyond the current state of natural language understanding. We present an approach that addresses these needs, using a large-scale knowledge base, simplified English grammar and a combination of compositional frame semantics and abductive reasoning. This in turn raises a significant challenge disambiguating complex semantic structures, which requires a pragmatics of narrative for constraint and guidance. We present a theory of narrative functions that serve as a heuristic for relevance in narrative, and provide evidence that this heuristic is effective for disambiguation that leads to consistent understanding.

6 citations

Proceedings ArticleDOI
11 Apr 2008
TL;DR: The approach to FIS is described which includes adversarial 'soft-factors' such as goals, rationale, and beliefs within a computational model that infers adversarial intent and allows the insertion of assumptions to be used in conjunction with current battlefield state to perform what-if analysis.
Abstract: Understanding the intent of today's enemy necessitates changes in intelligence collection, processing, and dissemination. Unlike cold war antagonists, today's enemies operate in small, agile, and distributed cells whose tactics do not map well to established doctrine. This has necessitated a proliferation of advanced sensor and intelligence gathering techniques at level 0 and level 1 of the Joint Directors of Laboratories fusion model. The challenge is in leveraging modeling and simulation to transform the vast amounts of level 0 and level 1 data into actionable intelligence at levels 2 and 3 that include adversarial intent. Currently, warfighters are flooded with information (facts/observables) regarding what the enemy is presently doing, but provided inadequate explanations of adversarial intent and they cannot simulate 'what-if' scenarios to increase their predictive situational awareness. The Fused Intent System (FIS) aims to address these deficiencies by providing an environment that answers 'what' the adversary is doing, 'why' they are doing it, and 'how' they will react to coalition actions. In this paper, we describe our approach to FIS which includes adversarial 'soft-factors' such as goals, rationale, and beliefs within a computational model that infers adversarial intent and allows the insertion of assumptions to be used in conjunction with current battlefield state to perform what-if analysis. Our approach combines ontological modeling for classification and Bayesian-based abductive reasoning for explanation and has broad applicability to the operational, training, and commercial gaming domains.

5 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