Tensor Representation of Topographically Organized Semantic Spaces

2018 ◽  
Vol 30 (12) ◽  
pp. 3259-3280
Author(s):  
Andrés Pomi ◽  
Eduardo Mizraji ◽  
Juan Lin
Keyword(s):  
Spatial Organization ◽  
Semantic Space ◽  
Tensor Representation ◽  
Host Matrix ◽  
Semantic Categories ◽  
Semantic Organization ◽  
Output Information ◽  
Experimental Findings ◽  
Human Brains ◽  
Flow Of Information

Human brains seem to represent categories of objects and actions as locations in a continuous semantic space across the cortical surface that reflects the similarity among categories. This vision of the semantic organization of information in the brain, suggested by recent experimental findings, is in harmony with the well-known topographically organized somatotopic, retinotopic, and tonotopic maps in the cerebral cortex. Here we show that these topographies can be operationally represented with context-dependent associative memories. In these models, the input vectors and, eventually also, the associated output vectors are multiplied by context vectors via the Kronecker tensor product, which allows a spatial organization of memories. Input and output tensor contexts localize matrices of semantic categories into a neural layer or slice and, at the same time, direct the flow of information arriving at the layer to a specific address, and then forward the output information toward the corresponding targets. Given a neural topographic pattern, the tensor representation will place a set of associative matrix memories within a topographic regionalized host matrix in such way that they reproduce the empirical pattern of patches in the actual neural layer. Progressive approximations to this goal are accomplished by avoiding excessive overlap of memories or the existence of empty regions within the host matrix.

A quantum-based semiotic model for textual semantics

Kybernetes ◽  
2018 ◽  
Vol 47 (2) ◽  
pp. 307-320 ◽  
Author(s):  
Francesco Galofaro ◽  
Zeno Toffano ◽  
Bich-Liên Doan
Keyword(s):  
Information Retrieval ◽  
Quantum Theory ◽  
Philosophy Of Language ◽  
Formal Model ◽  
Semantic Space ◽  
Turing Test ◽  
Point Of View ◽  
Content Type ◽  
Semantic Organization ◽  
Semiotic Model

Purpose The paper aims to provide a semiotic interpretation of the role played by entanglement in quantum-based models aimed to information retrieval and suggests possible improvements. Actual models are capable of retrieving documents relevant to a query composed of a keyword and its acceptation expressed by a given context. The paper also considers some analogies between this technique and quantum-based approaches in other disciplines to discuss the consequence of this quantum turn, as epistemology and philosophy of language are concerned. Design/methodology/approach We use quantum geometry to design a formal model for textual semiotics. In particular, the authors refer to Greimas’s work on semantics and information theory, to Eco’s writings on semantic memory and to Lotman’s work on a cybernetic notion of culture. Findings Quantum approaches imply a particular point of view on meaning. Meaning is not a real, positive quality of a given word. It is a net of relations constructed in the text, whose value is progressively determined during the reading process. Furthermore, reading is not a neutral operation: to read is to determine meaning. If it is said that, from a general semiotic point of view, meaning is stored in quantum semantic memories and is read/written by semantic machines, then the operation of “reading/writing” is analogous to the operation of measuring in quantum theory: in other terms, meaning is a value, and this implies an instance (not necessarily human) according to which values are valuable. Research limitations/implications The authors are not proposing a complete quantum semantics. At the present, quantum information retrieval can detect the presence of semantic relations. The authors suggest a way to characterize them, leaving open the problem on how to formalize the document as a vector in four-state semantic space. Practical implications A quantum turn shows deep semiotic implications on the approach to language, which shows an immanent semantic organization not reducible to syntax and morphology. This organization is probabilistic and indeterministic and explains to what extent text fixes the meaning of its lexical units. Social implications In the authors’ perspective, signification is not the exclusivity of a human subject. Criticizing Turing test, the great semiotic and cybernetic scholar Jurij Lotman wrote that if we identify “intelligent” and “human”, we raise the failings of an actual form of intelligence to the rank of an essential characteristic. On this line, meaning is considered as a feature of social, artificial and biological systems. Originality/value The adoption of quantum formalism seems in line with cybernetic framework, involving a probabilistic, non-cartesian point of view on meaning aimed to critically discuss the human–machine relation. Furthermore, Quantum theory (QT) implies a phenomenological point of view on the conditions of possibility of meaning.

An investigation of semantic space in patients with schizophrenia

1996 ◽  
Vol 2 (4) ◽  
pp. 267-273 ◽  
Author(s):  
Mark S. Aloia ◽  
Monica L. Gourovitch ◽  
Daniel R. Weinberger ◽  
Terry E. Goldberg
Keyword(s):  
Semantic Information ◽  
Semantic Networks ◽  
Semantic Space ◽  
Clustering Techniques ◽  
Word Generation ◽  
Semantic Organization ◽  
Specific Category ◽  
Relative Proximity ◽  
Schizophrenic Patients ◽  
Fluency Task

AbstractThere has been increasing interest in the semantic cognitive system in schizophrenia. Recent findings suggest a possible breakdown of semantic information processing in this disorder. The current study attempts to further examine semantic organization in schizophrenia. Twenty-eight chronic, early-onset schizophrenic patients and 32 controls were matched for prcmorbid intelligence and compared in their ability to spontaneously cluster exemplars from a specific category during a fluency task. Using multidimensional scaling and clustering techniques, 11 exemplars occurring most frequently in both groups were chosen for examination of their relative “proximity” during word generation. Patients with schizophrenia showed a less stable two-dimensional organization of exemplars and were less likely to group exemplars into subordinate clusters than were normals. These results suggest that semantic networks arc disorganized in these patients. These findings may have some implications for the debate over the origin of “thought disorder” in schizophrenia. (JINS, 1996, 2, 267–273.)

scholarly journals The Image of the Patient as a Component of the World’s Image of Students of Medical University

2020 ◽  
Vol 12 (1) ◽  
pp. 122-138
Author(s):  
N.V. Samoilenko ◽  
E.V. Dyachenko
Keyword(s):  
Medical Care ◽  
Clinical Teaching ◽  
Semantic Space ◽  
Professional Communication ◽  
Clinical Decisions ◽  
Educational Module ◽  
Semantic Categories ◽  
Methodological Foundation ◽  
The World ◽  
Medical University

The article provides data on the study of the patient’s image as a component of the image of the world among students of a medical university. The theoretical and methodological foundation was the provision that the image of the patient is implicitly represented in the semantic space of future doctors by semantic categories in the structure of a holistic multi-level system of ideas about the world, other people, about themselves and their activities, i.e. image of the world. The patient’s image was studied as a factor in the effectiveness of the interaction between the doctor and the patient in the provision of medical care. The effectiveness of interaction is one of the tools for making relevant clinical decisions, taking into account the opinion of the patient, which ensures patient-oriented medical consultation. 300 students of the 1 and 6 courses of the Ural State Medical University of the Faculty of Medicine were examined. The patient’s image was experimentally studied using methods: semantic personality differential, color metaphors and color choices. Based on the results obtained, it is concluded that the image of a patient in medical students is represented by various kinds of semantic categories (friendly, irrational, irritable, etc.). The content of the semantic categories of the patient’s image differs at the beginning and at the end of training: from multidirectional characteristics to ordered in the form of a typical and ideal patient. It has been established that the content of the patient’s image can be purposefully formed when mastering the educational program. The inclusion of an educational module in clinical teaching disciplines on the skills of professional communication with the patient (as one of the tools for making clinical decisions in the provision of medical care) affects the change in the content of the patient’s image in future doctors.

scholarly journals Predicting cognitive abilities across individuals using sparse EEG connectivity

2020 ◽  
Author(s):  
Nicole Hakim ◽  
Edward Awh ◽  
Edward K Vogel ◽  
Monica D Rosenberg
Keyword(s):  
Working Memory ◽  
Cognitive Abilities ◽  
Spatial Organization ◽  
Fluid Intelligence ◽  
Functional Organization ◽  
Working Memory Capacity ◽  
Neural Signals ◽  
Connectivity Patterns ◽  
Eeg Connectivity ◽  
Human Brains

ABSTRACTHuman brains share a broadly similar functional organization with consequential individual variation. This duality in brain function has primarily been observed when using techniques that consider the spatial organization of the brain, such as MRI. Here, we ask whether these common and unique signals of cognition are also present in temporally sensitive, but spatially insensitive, neural signals. To address this question, we compiled EEG data from individuals performing multiple working memory tasks at two different data-collection sites (ns = 171 and 165). Results revealed that EEG connectivity patterns were stable within individuals and unique across individuals. Furthermore, models based on these connectivity patterns generalized across datasets to predict participants’ working memory capacity and general fluid intelligence. Thus, EEG connectivity provides a signature of working memory and fluid intelligence in humans and a new framework for characterizing individual differences in cognitive abilities.

Who stole what from whom?

2019 ◽  
Vol 20 (1) ◽  
pp. 107-140
Author(s):  
Nicolás José Fernández-Martínez ◽  
Pamela Faber
Keyword(s):  
Machine Translation ◽  
Pattern Analysis ◽  
Word Sense Disambiguation ◽  
Semantic Space ◽  
Word Sense ◽  
Semantic Parsing ◽  
Frame Semantics ◽  
Semantic Categories ◽  
Sense Disambiguation ◽  
Grammar Model

Abstract Drawing on the Lexical Grammar Model, Frame Semantics and Corpus Pattern Analysis, we analyze and contrast verbs of stealing in English and Spanish from a lexico-semantic perspective. This involves looking at the lexical collocates and their corresponding semantic categories that fill the argument slots of verbs of stealing. Our corpus search is performed with the Word Sketch tool on Sketch Engine. To the best of our knowledge, no study has yet taken advantage of the Word Sketch tool in the study of the selection preferences of verbs of stealing, let alone a semantic, cross-linguistic study of those verbs. Our findings reveal that English and Spanish verbs of stealing map out the same underlying semantic space. This shared conceptual layer can thus be incorporated into an ontology based on deep semantics, which could in turn enhance NLP tasks such as word sense disambiguation, machine translation, semantic tagging, and semantic parsing.

scholarly journals A Generalized Language Model in Tensor Space

2019 ◽  
Vol 33 ◽  
pp. 7450-7458 ◽  
Author(s):  
Lipeng Zhang ◽  
Peng Zhang ◽  
Xindian Ma ◽  
Shuqin Gu ◽  
Zhan Su ◽  
...  
Keyword(s):  
Language Model ◽  
Tensor Decomposition ◽  
Expressive Power ◽  
Semantic Space ◽  
Language Models ◽  
Tensor Representation ◽  
Order Tensor ◽  
Tensor Space ◽  
Recursive Calculation ◽  
N Gram

In the literature, tensors have been effectively used for capturing the context information in language models. However, the existing methods usually adopt relatively-low order tensors, which have limited expressive power in modeling language. Developing a higher-order tensor representation is challenging, in terms of deriving an effective solution and showing its generality. In this paper, we propose a language model named Tensor Space Language Model (TSLM), by utilizing tensor networks and tensor decomposition. In TSLM, we build a high-dimensional semantic space constructed by the tensor product of word vectors. Theoretically, we prove that such tensor representation is a generalization of the n-gram language model. We further show that this high-order tensor representation can be decomposed to a recursive calculation of conditional probability for language modeling. The experimental results on Penn Tree Bank (PTB) dataset and WikiText benchmark demonstrate the effectiveness of TSLM.

scholarly journals PSYCHOSEMANTIC SPACE OF DIFFERENT LEVELS OF THE STUDENT YOUTH MORAL STABILITY

2019 ◽  
pp. 58-65
Keyword(s):  
Large Scale ◽  
Semantic Space ◽  
Basic Unit ◽  
Moral Consciousness ◽  
Consciousness Development ◽  
Semantic Organization ◽  
Individual System ◽  
The Individual ◽  
Different Levels ◽  
Rational Component

The article deals with theoretical-empirical measurements of the psycho-semantic space of different levels of moral consciousness of the individual as a specific, cultural and individual due to the construct of its integral psyche. The category of significance is substantiated theoretically as the basic unit of scientific study and analysis of the defined sphere. The denotative and connotative, as well as associative meaning as the embodiment of the unity of a rational component in any sense with its rational, intuitive-sensual foundation, are described. As a result of the systematic theoretical and empirical study of the problem, the psycho-semantic organization of moral consciousness is presented through a description of its psycho-semantic space. The psycho-semantic space of the moral consciousness of the student's personality is defined as an individual system of values in the totality of the corresponding individual and cultural heritage of the individual, structured in the sensual fabric of consciousness in the form of associative meanings of the notion of morality, the most frequent of which is the "conscience." Also, the connotative meanings of the basic notion of conscience on the three main levels of moral consciousness development in student youth were identified and interpreted: with leading evaluation and dynamically active connotations – at the preconventional level, with strong and ambivalent evaluative connotations – at the conventional level, positively evaluated and large-scale connotations – at the post-conventional level.

scholarly journals THE SEMANTIC ORGANIZATION OF DIPLOMATIC TERMS

2020 ◽  
Vol 72 (2) ◽  
pp. 152-162
Author(s):  
K. Ashinova ◽  
Keyword(s):  
Word Meaning ◽  
Semantic Content ◽  
Word Formation ◽  
Semantic Interpretation ◽  
Semantic Categories ◽  
Single Root ◽  
Semantic Organization ◽  
General Rules ◽  
Complex Term ◽  
Main Thing

This article provides a brief description of semantic processes such as expansion, contraction or specialization of meanings. Nowadays there is no common understanding and continuity of main semantic categories known in the field of linguistics. It is known such categories as sign (symbol), concept, meaning, types of linguistic meanings, absolute and relative semantic content and others are included in that.The main thing in terminology process is word meaning, and it is determined by main featuresof term concept. As a result of semantic development of ancient words and changes in internal semantic structure the new nominative semewill be defined. Language is directly related to processes of differentiation and integration. These processes are characterized by semantic interpretation of language signs in understanding and communication. The diplomatic terms are formed according to general rules of word formation striving for individuality and stylistic neutrality. The meaning of complex term does not derive from individual meaning of its components. The component function is equal to function of phonemes in single-root terms particularly in distinctive form. The article was prepared on the basis of written sources and literature.

scholarly journals Semantic Categories in the Domain of Motion Verbs by Adult Speakers of Danish, German, and Turkish

2014 ◽  
Vol 61 (4) ◽  
Author(s):  
Moiken Jessen
Keyword(s):  
Semantic Space ◽  
Motion Verbs ◽  
Motion Events ◽  
Fine Grained ◽  
Semantic Categories ◽  
Video Clips ◽  
The World ◽  
Motion Verb ◽  
Vector Orientation ◽  
Verb Lexicon

Languages differ in the ways they divide the world. This study applies cluster analysis to understand how and why languages differ in the way they express motion events. It further lays out what the parameters of the structure of the semantic space of motion are, based on data collected from participants who were adult speakers of Danish, German, and Turkish. The participants described 37 video clips depicting a large variety of motion events. The results of the study show that the segmentation of the semantic space displays a great deal of variation across all three groups. Turkish differs from German and Danish with respect to the features used to segment the semantic space – namely by using vector orientation. German and Danish differ greatly with respect to (a) how fine-grained the distinctions made are, and (b) how motion verbs with a common Germanic root are distributed across the semantic space. Consequently, this study illustrates that the parameters applied for categorization by speakers are, to some degree, related to typological membership, in relation to Talmy's typological framework for the expression of motion events. Finally, the study shows that the features applied for categorization differ across languages and that typological membership is not necessarily a predictor of elaboration of the motion verb lexicon.

Sign in / Sign up

Export Citation Format

Share Document