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.

A kind of mixed reasoning in rule-based systems

Abstract: The method of mixed reasoning is put forward in this paper. It can accomplish deductive and abductive reasoning alternatively in a problem-solving process. A problem solving model of the method is defined. The logic reasoning in this model can achieve a change of the reasoning direction. In this paper, some other problems and technologies of completing mixed reasoning are also discussed, such as how the reasoning algorithm is designed and its properties are analysed.

Finding Evidence of Deductive and Abductive Reasoning Abilities: An Experiment with Lateral-Brain-Damage Patients

Abstract: Finding Evidence of Deductive and Abductive Reasoning Abilities: An Experiment with Lateral-Brain-Damage Patients Franco Salvetti (franco.salvetti@colorado.edu) Dept. of Computer Science University of Colorado at Boulder 430 UCB, Boulder, CO 80309–0430, U.S.A. Nicolas Nicolov (nicolas@umbrialistens.com) Umbria Inc. 1655 Walnut Str. Boulder, CO 80302, U.S.A. Abstract as associated with the right hemisphere (Code, 1987). By binding the concept of logical deduction in reasoning with prediction of the future, it appears that patients with left hemi- sphere damage, on the other hand, linguistically attend more to prediction than explanation, even though linguistic evi- dence of both types of reasoning is less than among members of the control group. The implications of this evidence and possible linguistic and cognitive explanations for this evidence are discussed. In narrative discourse, explanation can be considered abduc- tive reasoning, described as a creative type of reasoning which generates new ideas, and prediction can be considered logical deductive reasoning. Given the well-known associations with hemisphere processing and reasoning associations, it would be expected that patients with left hemisphere brain damage would exhibit more evidence of explanation versus prediction in reasoning and patients with right hemisphere brain damage would exhibit more evidence of prediction versus explanation in reasoning. By observing actual linguistic evidence of rea- soning in left and right hemisphere brain damaged patients in comparison to that of a control group, we were able to dis- cover important differences between these groups. Contrary to assumed wisdom, patients with left hemisphere damage pro- duced more utterances of prediction than of explanation and patients with right hemisphere damage produced more utter- ances of explanation than prediction on a narrative discourse task. Linguistic processing and cognitive processing are dis- cussed as explanations for these results. Related Research Several previous studies, e.g., (Geminiani & Bucciarelli, 1998), have addressed the effect of lateral brain damage on explanation and prediction reasoning processes. By and large, the left hemisphere has been more commonly asso- ciated with language processing and the right hemisphere with visuospatial processing and abstract reasoning. Re- lated studies, e.g., (Sacco, Bucciarelli, & Adenzato, 2001), have focused on causal reasoning, theory of mind processing, metaphor interpretation, and discourse abilities. Mind processing and simple causal reasoning were shown to be intact in a severe agrammatic aphasic patient with a large lesion in the left hemisphere who could not formu- late propositions in speech or writing, make judgments as to whether a sentence is grammatical, match sentences to pic- tures, or identify the meaning of verbs. This patient tested in the 91st percentile on the Wisconsin card sorting test and was able to rely on the visuospatial representation and memory of the location and attributes of objects in order to communicate his responses. Grammar may therefore play a vital role in the development of cognitive processes, but once these processes are established, cognition can operate without grammar (Sie- gal, Varley, & Want, 2001). The assumption that reasoning based on world knowledge is intact in severe aphasia may be questioned (Joanette & Brownell, 1990). Huber discusses conflicting evidence that Global and Wernicke’s aphasic patients had as much difficul- ties ordering picture stories as did right hemisphere patients. Not only may visuospatial reasoning be damaged, but general impairment of reasoning or sequencing may be attributed to these common deficits. Huber (Joanette & Brownell, 1990) found that choosing a figurative meaning of an idiom is more demanding than choosing the literal meaning for normals, pa- tients with right hemisphere damage, and aphasic patients. Idioms with close relationships between the literal and figura- Introduction Human beings are able to generate long chains of reasoning which produce inferences (Birnbaum, 1986). A way to cat- egorize different kinds of reasoning is to focus on reasoning which produces prediction versus reasoning which explains observations. These two tasks are here considered cognitively different. The first is considered more closely related to clas- sical logical deduction (Feferman, 1999), while the second to abductive reasoning (Magnani, 2005) and creative thinking. Analyzing transcripts from an experiment, administered at Boston VA Hospital and fully described here, that involved patients with lateral brain damage, we have found prelimi- nary evidence that patients affected by aphasia exhibit ver- bal evidence of reasoning which, compared with a control group, suggests that while capable of performing reasoning, they have less problems in producing sentences with evidence of prediction than sentences with evidence of explanation. Patients with right–hemisphere brain damage exhibit verbal evidence of reasoning which, when compared with a control group, suggests that even though abstract reasoning and the ability to make inferences may be impaired, they have fewer problems producing sentences with evidence of explanation than sentences with evidence of prediction. By binding the concept of creativity in reasoning with ex- planation in the past, it appears that patients with right hemi- sphere brain damage attend even more to this type of reason- ing than normals, even though creativity could be thought of

Reframing the Problematic: From Translation to the Interaction of Theory to Practice in Teacher Education

Abstract: The translation of theory to practice is a long-standing ‘wicked’ problem in initial teacher education. Recent policy debates on this problem have focused on superficial issues like the location of the program in either school or universities. These debates have overlooked the critical issue of the mechanisms within these programs that promote or hinder the translation of theory to practice for their students. The authors believe that the critical policy debate establishes a false binary that is not helpful to the practice of teacher education. This book has presented an empirically rigorous and theoretically robust discussion about what enhances the translation of theory to practice in initial teacher education regardless of location. This discussion has highlighted the important of the fundamental relationship between effective translation and implementation science. This relationship is underpinned by a developmental model of learning characterised by reciprocity, collaboration and shared goals between all the stakeholders in initial teacher education. Translation is also enhanced through models of teacher education that grant equal status to practice and theory. This equality negates the need for translation as the program logic and design embodies the abductive reasoning required when learning how to teach.

Generating Hypothetical Events for Abductive Inference

Abstract: Abductive reasoning starts from some observations and aims at finding the most plausible explanation for these observations. To perform abduction, humans often make use of temporal and causal inferences, and knowledge about how some hypothetical situation can result in different outcomes. This work offers the first study of how such knowledge impacts the Abductive NLI task -- which consists in choosing the more likely explanation for given observations. We train a specialized language model LMI that is tasked to generate what could happen next from a hypothetical scenario that evolves from a given event. We then propose a multi-task model MTL to solve the Abductive NLI task, which predicts a plausible explanation by a) considering different possible events emerging from candidate hypotheses -- events generated by LMI -- and b) selecting the one that is most similar to the observed outcome. We show that our MTL model improves over prior vanilla pre-trained LMs fine-tuned on Abductive NLI. Our manual evaluation and analysis suggest that learning about possible next events from different hypothetical scenarios supports abductive inference.

Through the Eyes of an Archeologist: Studying the Role of Prior Knowledge in Learning with Diagrams.

Abstract: The study of learning with diagrams predominantly embraces a cognitivist viewpoint in which learning and problem solving are conceptualized as interplay between internal mental and external physical representations. We start by questioning two of its underlying assumptions originating in objectivist epistemology: the directness of diagrams as iconic representations and the sameness, in terms of cultural and linguistic background, of the producers and users of diagrams. In the remaining of the contribution, a relativistic viewpoint is adopted according to which reality is socially and experientially based and representation crucially depends on an observer within a context. Consequently, we investigate a theoretical situation in which learners have knowledge of neither the content nor the type of diagram. As a method of study, we adopt the perspective of an archeologist who encounters inscriptions of unknown civilizations, and in doing so, we uncover the role of abductive reasoning in learning with diagrams. A formal framework is developed with recursive rules for further studying the distinctions between pictures and diagrams, literal (denotative), figurative (connotative), and metalinguistic meaning, as well as internal and external representations. The framework is illustrated with a set of content-free test patterns. Some directions for evaluating the framework are presented in the conclusion.