Representation Of Semantic Concepts Research Articles

Sensorimotor norms for Chinese nouns and their relationship with orthographic and semantic variables

ABSTRACT Sensorimotor information is vital to the conceptual representation of our knowledge system. This study collects perceptual and action ratings for 664 disyllabic nouns among 438 native speakers and creates the first and largest dataset of sensorimotor norms for nouns in Chinese. Using aggregated semantic covariates, including concreteness ratings from a concreteness rating study, as well as the reaction times and error rates from a lexical decision study, our current work demonstrates the strengths of sensory modalities and action effectors in Chinese nouns and explores the contributions of embodied experiences in reflecting orthographic representations and semantic processing in the Chinese language. This study contributes valuable data sources to the study of Chinese lexical processing and highlights the importance of sensorimotor information and embodied manifestations in the semantic representations of concepts. Our results also support the language universal that orthographic awareness in lexical processing and reading supersedes phonological awareness.

Evidence of semantic processing impairments in behavioural variant frontotemporal dementia and Parkinson's disease.

Category-specific impairments caused by brain damage can provide important insights into how semantic concepts are organized in the brain. Recent research has demonstrated that disease to sensory and motor cortices can impair perceptual feature knowledge important to the representation of semantic concepts. This evidence supports the grounded cognition theory of semantics, the view that lexical knowledge is partially grounded in perceptual experience and that sensory and motor regions support semantic representations. Less well understood, however, is how heteromodal semantic hubs work to integrate and process semantic information. Although the majority of semantic research to date has focused on how sensory cortical areas are important for the representation of semantic features, new research explores how semantic memory is affected by neurodegeneration in regions important for semantic processing. Here, we review studies that demonstrate impairments to abstract noun knowledge in behavioural variant frontotemporal degeneration (bvFTD) and to action verb knowledge in Parkinson's disease, and discuss how these deficits relate to disease of the semantic selection network. Findings demonstrate that semantic selection processes are supported by the left inferior frontal gyrus (LIFG) and basal ganglia, and that disease to these regions in bvFTD and Parkinson's disease can lead to categorical impairments for abstract nouns and action verbs, respectively.

Human Cortical Neurons in the Anterior Temporal Lobe Reinstate Spiking Activity during Verbal Memory Retrieval

When we recall an experience, we rely upon the associations that we formed during the experience, such as those among objects, time, and place [1]. These associations are better remembered when they are familiar and draw upon generalized knowledge, suggesting that we use semantic memory in the service of episodic memory [2, 3]. Moreover, converging evidence suggests that episodic memory retrieval involves the reinstatement of neural activity that was present when we first experienced the event. Therefore, we hypothesized that retrieving associations should also reinstate the neural activity responsible for semantic processing. Indeed, previous studies have suggested that verbal memory retrieval leads to the reinstatement of activity across regions of the brain that include the distributed semantic processing network [4-6], but it is unknown whether and how individual neurons in the human cortex participate in the reinstatement of semantic representations. Recent advances using high-density microelectrode arrays (MEAs) have allowed clinicians to record from populations of neurons in the human cortex [7, 8]. Here we used MEAs to record neuronal spiking activity in the human middle temporal gyrus (MTG), a cortical region supporting the semantic representation of words [9-11], as participants performed a verbal paired-associates task. We provide novel evidence that population spiking activity in the MTG forms distinct representations of semantic concepts and that these representations are reinstated during the retrieval of those words.

Characterizing everyday activities from visual lifelogs based on enhancing concept representation

The proliferation of wearable visual recording devices such as SenseCam, Google Glass, etc. is creating opportunities for automatic analysis and usage of digitally-recorded everyday behavior, known as visual lifelogs. Such information can be recorded in order to identify human activities and build applications that support assistive living and enhance the human experience. Although the automatic detection of semantic concepts from images within a single, narrow, domain has now reached a usable performance level, in visual lifelogging a wide range of everyday concepts are captured by the imagery which vary enormously from one subject to another. This challenges the performance of automatic concept detection and the identification of human activities because visual lifelogs will have such variety of semantic concepts across individual subjects. In this paper, we characterize the everyday activities and behavior of subjects by applying a hidden conditional random field (HCRF) algorithm on an enhanced representation of semantic concepts appearing in visual lifelogs. This is carried out by first extracting latent features of concept occurrences based on weighted non-negative tensor factorization (WNTF) to exploit temporal patterns of concept occurrence. These results are then input to an HCRF-based model to provide an automatic annotation of activity sequences from a visual lifelog. Results for this are demonstrated in experiments to show the efficacy of our algorithm in improving the accuracy of characterizing everyday activities from individual lifelogs. The overall contribution is a demonstration that using images taken by wearable cameras we can capture and characterize everyday behavior with a level of accuracy that allows useful applications which measure, or change that behavior, to be developed.

An unsupervised method for learning generation dictionaries for spoken dialogue systems by mining user reviews

Spoken language generation for dialogue systems requires a dictionary of mappings between the semantic representations of concepts that the system wants to express and the realizations of those concepts. Dictionary creation is a costly process; it is currently done by hand for each dialogue domain. We propose a novel unsupervised method for learning such mappings from user reviews in the target domain and test it in the restaurant and hotel domains. Experimental results show that the acquired mappings achieve high consistency between the semantic representation and the realization and that the naturalness of the realization is significantly higher than the baseline.

Determining the dimensionality in spatial representations of semantic concepts

When multidimensional scaling solutions are used to study semantic concepts, the dimensionality of the optimal configuration has to be determined. Several strategies have been proposed to choose the appropriate dimensionality. In the present paper, the traditional dimensionality choice criteria were evaluated and compared to a method based on the prediction of an external criterion. Two studies were conducted in which typicality of an exemplar within a semantic concept was predicted from its distance to the concept centroid. In contrast to the low-dimensional solutions selected by the traditional methods, predictions of an external criterion improved with additional dimensions up till dimensionalities that were much higher than what is common in the literature. This suggests that traditional methods underestimate the richness of semantic concepts as revealed in spatial representations derived from similarity measures.

Analysis of the semantic representations of living and nonliving concepts: A normative study

Current feature-based semantic memory models assume that the semantic representations of concepts differ systematically across living and nonliving categories and that such differences account for the emergence of category-specific semantic deficits in brain-damaged people. To assess some of the different models' main assumptions about structural differences at the semantic feature level in the two major semantic domains, we administrated a feature-listing task to normal young volunteers on 64 concepts drawn from living and nonliving semantic categories. We investigated whether feature correlation, a variable with a crucial role in the emergence of category-specific deficits, should be computed as a concept-dependent or as a concept-independent measure, and we chose the former. We also addressed the issue of a psychological counterpart of feature production frequency. Finally, we analysed the database obtained from the feature-listing tasks, looked at cross-domain differences for correlation, feature frequency, distinctiveness, and feature type, and discussed the implications of these findings for current semantic memory models.

Computer Aided Cognition to Support Problem-Centered Decomposition of Complex Problems

Complex problems such as analysis of military situation assessment, homeland defense, diagnosis of the health of complex systems, medical diagnosis, and environmental monitoring require the ability to utilize a wide variety of data such as signals, images, textual information, and scalar data. The rapid evolution of micro-scale sensors, wideband communications, and microprocessors enables the collection and dissemination of huge amounts of data to be provided to a human analyst. Unfortunately, the analyst cannot directly understand nor process the data. Instead, analysts reason about high-level abstractions via language. A challenge exists to decompose general problems into detailed models that link to specific types of data (viz., problem centered decomposition) and to compose data into meaningful relationships to assist the understanding of semantic representations of abstract concepts. This paper discusses the challenge of problem-centered analysis (including problem centered decomposition and problem centered composition) and describes our efforts to develop cognitive aids to assist the analysis process for improved understanding of complex problems.