Distributed Code for Semantic Relations Predicts Neural Similarity during Analogical Reasoning

2020 ◽  
pp. 1-13 ◽  
Author(s):  
Jeffrey N. Chiang ◽  
Yujia Peng ◽  
Hongjing Lu ◽  
Keith J. Holyoak ◽  
Martin M. Monti

The ability to generate and process semantic relations is central to many aspects of human cognition. Theorists have long debated whether such relations are coarsely coded as links in a semantic network or finely coded as distributed patterns over some core set of abstract relations. The form and content of the conceptual and neural representations of semantic relations are yet to be empirically established. Using sequential presentation of verbal analogies, we compared neural activities in making analogy judgments with predictions derived from alternative computational models of relational dissimilarity to adjudicate among rival accounts of how semantic relations are coded and compared in the brain. We found that a frontoparietal network encodes the three relation types included in the design. A computational model based on semantic relations coded as distributed representations over a pool of abstract relations predicted neural activities for individual relations within the left superior parietal cortex and for second-order comparisons of relations within a broader left-lateralized network.

2019 ◽  
Author(s):  
Jeffrey N. Chiang ◽  
Yujia Peng ◽  
Hongjing Lu ◽  
Keith J. Holyoak ◽  
Martin M. Monti

AbstractThe ability to generate and process semantic relations is central to many aspects of human cognition. Theorists have long debated whether such relations are coded as atomistic links in a semantic network, or as distributed patterns over some core set of abstract relations. The form and content of the conceptual and neural representations of semantic relations remains to be empirically established. The present study combined computational modeling and neuroimaging to investigate the representation and comparison of abstract semantic relations in the brain. By using sequential presentation of verbal analogies, we decoupled the neural activity associated with encoding the representation of the first-order semantic relation between words in a pair from that associated with the second-order comparison of two relations. We tested alternative computational models of relational similarity in order to distinguish between rival accounts of how semantic relations are coded and compared in the brain. Analyses of neural similarity patterns supported the hypothesis that semantic relations are coded, in the parietal cortex, as distributed representations over a pool of abstract relations specified in a theory-based taxonomy. These representations, in turn, provide the immediate inputs to the process of analogical comparison, which draws on a broad frontoparietal network. This study sheds light not only on the form of relation representations but also on their specific content.SignificanceRelations provide basic building blocks for language and thought. For the past half century, cognitive scientists exploring human semantic memory have sought to identify the code for relations. In a neuroimaging paradigm, we tested alternative computational models of relation processing that predict patterns of neural similarity during distinct phases of analogical reasoning. The findings allowed us to draw inferences not only about the form of relation representations, but also about their specific content. The core of these distributed representations is based on a relatively small number of abstract relation types specified in a theory-based taxonomy. This study helps to resolve a longstanding debate concerning the nature of the conceptual and neural code for semantic relations in the mind and brain.


Author(s):  
Kim Uittenhove ◽  
Patrick Lemaire

In two experiments, we tested the hypothesis that strategy performance on a given trial is influenced by the difficulty of the strategy executed on the immediately preceding trial, an effect that we call strategy sequential difficulty effect. Participants’ task was to provide approximate sums to two-digit addition problems by using cued rounding strategies. Results showed that performance was poorer after a difficult strategy than after an easy strategy. Our results have important theoretical and empirical implications for computational models of strategy choices and for furthering our understanding of strategic variations in arithmetic as well as in human cognition in general.


Author(s):  
Yu.V. Kupriyanova ◽  
I.M. Vasilyanova

The article summarizes the key points in the development of the metadialogue phenomenon from a linguistic point of view. Some stages of the development of this concept and the difficulties associated with its structuring are covered. The main research findings of modern foreign and domestic experts on its study are considered. Some characteristics of the subject of the research from the standpoint of various pragmatic installations are given. On the basis of the dynamic structure of the metadialogue development, certain principles of semantic relations connected with the dialectical nature of human cognition are presented. Excursion into the history and evolution of the concept is presented. Several types of formulation of the subject matter are given. In accordance with the goal of speech exposure, internal problems of the development of metadialogue are highlighted and the critical points related to solving these problems are described. The rules of metadialogue flow are explained at the level of steps, the success/failure of which directly affects the final result of communication. The prospects of development of the concept research in accordance with various types of discourse are indicated.


2019 ◽  
Author(s):  
Allison Letkiewicz ◽  
Amy L. Cochran ◽  
Josh M. Cisler

Trauma and trauma-related disorders are characterized by altered learning styles. Two learning processes that have been delineated using computational modeling are model-free and model-based reinforcement learning (RL), characterized by trial and error and goal-driven, rule-based learning, respectively. Prior research suggests that model-free RL is disrupted among individuals with a history of assaultive trauma and may contribute to altered fear responding. Currently, it is unclear whether model-based RL, which involves building abstract and nuanced representations of stimulus-outcome relationships to prospectively predict action-related outcomes, is also impaired among individuals who have experienced trauma. The present study sought to test the hypothesis of impaired model-based RL among adolescent females exposed to assaultive trauma. Participants (n=60) completed a three-arm bandit RL task during fMRI acquisition. Two computational models compared the degree to which each participant’s task behavior fit the use of a model-free versus model-based RL strategy. Overall, a greater portion of participants’ behavior was better captured by the model-based than model-free RL model. Although assaultive trauma did not predict learning strategy use, greater sexual abuse severity predicted less use of model-based compared to model-free RL. Additionally, severe sexual abuse predicted less left frontoparietal network encoding of model-based RL updates, which was not accounted for by PTSD. Given the significant impact that sexual trauma has on mental health and other aspects of functioning, it is plausible that altered model-based RL is an important route through which clinical impairment emerges.


2020 ◽  
Author(s):  
Svetla Koeva ◽  
Svetlozara Leseva ◽  
Ivelina Stoyanova ◽  
Maria Todorova ◽  
Hristina Kukova ◽  
...  

2012 ◽  
Vol 367 (1585) ◽  
pp. 103-117 ◽  
Author(s):  
Katerina Pastra ◽  
Yiannis Aloimonos

Language and action have been found to share a common neural basis and in particular a common ‘syntax’, an analogous hierarchical and compositional organization. While language structure analysis has led to the formulation of different grammatical formalisms and associated discriminative or generative computational models, the structure of action is still elusive and so are the related computational models. However, structuring action has important implications on action learning and generalization, in both human cognition research and computation. In this study, we present a biologically inspired generative grammar of action, which employs the structure-building operations and principles of Chomsky's Minimalist Programme as a reference model. In this grammar, action terminals combine hierarchically into temporal sequences of actions of increasing complexity; the actions are bound with the involved tools and affected objects and are governed by certain goals. We show, how the tool role and the affected-object role of an entity within an action drives the derivation of the action syntax in this grammar and controls recursion, merge and move, the latter being mechanisms that manifest themselves not only in human language, but in human action too.


Author(s):  
Michael N. Jones ◽  
Jon Willits ◽  
Simon Dennis

Meaning is a fundamental component of nearly all aspects of human cognition, but formal models of semantic memory have classically lagged behind many other areas of cognition. However, computational models of semantic memory have seen a surge of progress in the last two decades, advancing our knowledge of how meaning is constructed from experience, how knowledge is represented and used, and what processes are likely to be culprit in disorders characterized by semantic impairment. This chapter provides an overview of several recent clusters of models and trends in the literature, including modern connectionist and distributional models of semantic memory, and contemporary advances in grounding semantic models with perceptual information and models of compositional semantics. Several common lessons have emerged from both the connectionist and distributional literatures, and we attempt to synthesize these themes to better focus future developments in semantic modeling.


2019 ◽  
Vol 6 (2) ◽  
pp. 415-429
Author(s):  
Fang Wang ◽  
Lijun Lu ◽  
Lu Xu ◽  
Bihu Wu ◽  
Ying Wu

Purpose Tourists’ destination image is crucial for visiting intentions. An ancient capital with diverse characteristics is an important component of China’s urban tourism. The purpose of this paper is to address the following questions: what are the differences and commonalities of the perceived destination image of ancient capitals? What makes the difference of the perceived destination image in these cities? Aside from the exterior factors, are there internal factors of cities that influence tourists’ cognition and perception of destination image? Design/methodology/approach The comment text data of Baidu tourism website were used to determine the differences in the destination images of China’s four great ancient capitals: Beijing, Xi’an, Nanjing and Luoyang. ROST content mining and semantic network analysis were for differences and commonalities of the perceived destination image, and correlation analysis was used to explore the internal factors of cities that influence tourists’ cognition and perception of destination image. Findings Though the same as ancient capital, the four ancient capitals’ images are far apart; historical interests are the core of tourism experience in ancient capital city; image perception is from physical carrier, history and culture, and human cognition; tourist’ destination affect of ancient capital is most from its history and culture; protecting identity and maintaining daily life are crucial for ancient city tourism. Originality/value Previous studies on ancient capitals have focused on the invariable identity of ancient capitals’ destination images, and left a gap on determining from where the invariable identity comes in general and how much it influences destination image. This gap was addressed in this study, by analyzing the destination images of four ancient capitals in China as cases. In this way, this study provided reference to the other ancient cities worldwide.


2021 ◽  
Author(s):  
Isaac Treves

Prediction is a fundamental process in human cognition. Prediction means extracting one or more statistics from the distribution of past inputs and using that information to make a decision. What are the statistics underlying human predictions, and how do they change with training? To investigate these questions, we designed a sequence termination task, where participants watch temporally unfolding sequences and terminate them when they can predict the next item. We then test how well the participants’ termination points are predicted by computational models. We contrast frequency estimation models (How often did this symbol appear in the sequence?), transition models (How often did symbol A follow symbol B?), and a chunking model (What are the patterns of symbols?). In an online experiment with 65 adults, we find that participants are best fit by a transition-counting model. To assess the effect of training, we manipulated passive exposure to the sequences prior to the sequence termination task. Contrary to our expectations, prior exposure to sequences had no effect on termination performance– whether tested statistically or computationally, and despite good power. Lastly, training specifically on the termination task may shift responses towards chunking. These results provide insight into the representations, or information in mind, behind prediction. However, the lack of an effect of prior exposure makes it clear that sequence termination measures explicit, or conscious, prediction. Future work could examine whether representations in explicit prediction tasks like sequence termination are different from implicit, or unconscious, tasks like the serial reaction time task.


Author(s):  
Roozbeh Sanaei ◽  
Wei Lu ◽  
Luciënne T. M. Blessing ◽  
Kevin N. Otto ◽  
Kristin L. Wood

Analogy-making has been deemed one of the core cognitive mechanisms which play a role in human creative thinking activities such as design and art. Designers can make use of analogies in various stages of design including ideation, planning and evaluation. However, human analogy-making is limited by experience and reliance of human memory on superficial attributes rather than relational or causal structure during analogy retrieval. In this regard, different design-by-analogy tools have been developed to assist designers in analogical reasoning. Analogical reasoning tools can be viewed as either based on hand-coded structured knowledge or natural-language-based design-by-analogy tools. The former are naturally limited in extent and scope to that which was hand coded [1]. Alternatively, natural language analogical reasoning can leverage the abundantly available textual resources. Current text-based analogy research for design have relied on analogies between individual word meanings. This leaves open consideration of the relational structure of the language where the relational similarity of texts can indicate a significant analogy. In this article, we develop four computational models of analogy that capture relational structure of the text. This includes spatial representation of semantics, multi-level deep neural reasoning, graph matching based model and transformation-based model. The models are then combined together into an ensemble model to achieve acceptable level of analogical accuracy for the end-user. The underlying design-related knowledge upon which analogies were drawn includes engineering ontologies, function hierarchy and raw patent texts. Instantiating this analogical reasoning model in design concept analogy retrieval system, we show this approach can help retrieve meaningful analogies from the World Intellectual Property Organization (WIPO) patent repository. We demonstrate this for a particular design problem.


Sign in / Sign up

Export Citation Format

Share Document