WAYS OF ACTUALIZATION OF NATIONAL PERCEPTION OBJECTS OF ARCHITECTURE

Problem statement. Today in Ukraine the mechanisms of public influence on decision-making regarding the development of urban ensembles and cities in general continue to be formed. At the same time, the level of education of ordinary citizens in current trends in architecture, design and urban planning remains quite low. This affects the general semantic state of the architectural content of cities and results in a literal repetition of the visual forms of the XIX century in modern buildings, especially in the central (historical) parts of Ukrainian cities. According to the hypothesis of this research, the solution to this problem is in the creation of a system of free basic architectural education at several different levels. Purpose of the article. The main purpose of this work is to determine ways to update the perception of architectural objects by residents of Ukrainian cities. The objectives of the study are the analysis of publicly available methods of obtaining basic architectural and urban education and their classification, the definition of rational means to increase the general level of awareness in the field of architecture and urban planning. Conclusion. Due to the extraordinary breadth and heterogeneity of the population of large cities of Ukraine, achieving a global departure from inauthentic and irrelevant architectural and urban principles and update the national perception of architectural objects and architecture in general is possible only through introducing a set of actions aimed at a specific target audience (TA). In this study, in order to simplify, only four basic TAs are identified, but in further developments they need to be detailed more to achieve a better actual result. In the general context, the most important will be TA1 (students) − an audience that can begin to receive basic architectural knowledge, sufficient enough to understand the main trends in architecture and urban planning.

Real-Time Semantic Segmentation with Dual Encoder and Self-Attention Mechanism for Autonomous Driving

As the techniques of autonomous driving become increasingly valued and universal, real-time semantic segmentation has become very popular and challenging in the field of deep learning and computer vision in recent years. However, in order to apply the deep learning model to edge devices accompanying sensors on vehicles, we need to design a structure that has the best trade-off between accuracy and inference time. In previous works, several methods sacrificed accuracy to obtain a faster inference time, while others aimed to find the best accuracy under the condition of real time. Nevertheless, the accuracies of previous real-time semantic segmentation methods still have a large gap compared to general semantic segmentation methods. As a result, we propose a network architecture based on a dual encoder and a self-attention mechanism. Compared with preceding works, we achieved a 78.6% mIoU with a speed of 39.4 FPS with a 1024 × 2048 resolution on a Cityscapes test submission.

An Improved TextRank Multi-feature Fusion Algorithm For Keyword Extraction of Educational Resources

In view of the fact that the traditional graph model method which only considers statistical features or general semantic features when extracting keywords from existing massive educational resources, lacks the function of mining and utilizing multi-factor semantic features, this paper proposes an improved TextRank-based algorithm for keyword extraction of educational resources. According to the characteristics of Chinese text and the shortcomings of traditional TextRank algorithm, the improved algorithm featuring multi-feature fusion is developed using the importance of words in the corpus, the location information in the text and the attributes of words. Experimental results show that this method has higher accuracy, recall rate, and F-measure value than traditional algorithms in the process of keyword extraction of educational resources, which improves the quality of keyword extraction and is beneficial to better utilization and management of educational resources.

Visions of lexicography of a semantic European

In this essay, it is assumed that the languages of Latin Europe do have many semantic features in common, which contradicts the prevailing view of a general semantic particularity of every individual language and thus the exploitation for national-political purposes arising from that view. However, the proposition made here requires a summary and the assessment of different semantic concepts led by the idea of commonality. By means of individual cases that can be understood as relevant examples, a vision of lexicography will follow that aims at replacing the biologistic concept of a genetic explanation for contrastive semantics by the concept of a comparative semantics that is based on socio-historical, cultural-historcial and textual-historical arguments. In doing so, a historiography relating to the subject-matter of “semantics” will be suggested that assigns a semantic bridging function to Late Antiquity / Early Medieval Latin in relation to all languages of Latin Europe. The logic of the argument implies that a new era of semantic history begins upon the development of a structure of national languages in Europe, whose historical basis can still be recognised in the semantic communalities.

Coexistence of the Social Semantic Effect and Non-Semantic Effect in the Default Mode Network

Neuroimaging studies have found both semantic and non-semantic effects in the default mode network (DMN), leading to an intense debate on the role of the DMN in semantic processes. Four different views have been proposed: 1) The general semantic view holds that the DMN contains several hub regions supporting general semantic processes; 2) the non-semantic view holds that the semantic effects observed in most regions of the DMN (especially the ventral angular gyrus) are confounded by difficulty and do not reflect semantic processing per se; 3) the multifunction view holds that the same areas in the DMN can support both semantic and non-semantic functions; and 4) the multisystem view holds that the DMN contains multiple subnetworks supporting different aspects of semantic processes separately. Using an fMRI experiment, we found that in one of the subnetworks of the DMN, called the social semantic network, all areas showed social semantic activation and difficulty-induced deactivation. The ventral angular gyrus, whose function had been interpreted according to the difficulty effect, showed social semantic activation independent of difficulty. In addition, the distributions of two non-semantic effects, that is, difficulty-induced and task-induced deactivations, showed dissociation in the DMN. Our findings provide two insights into the semantic and non-semantic functions of the DMN, which are consistent with both the multisystem and multifunction views: First, the same areas of the DMN can support both social semantic and non-semantic functions; second, similar to the multiple semantic effects of the DMN, the non-semantic effects also vary across its subsystems.

For the Decoding of a Proverb in a Speech Act

As is known a proverb (both formally and semantically) is the micro-model of the life and mentality of the world where it was created and where it operates. Its basic function is grasping the wisdom seen from the viewpoint of the given ethnos – the general regularities. It is figuratively constructed of the specific material which is recorded in “the sensory material of perception” of the given ethnos, i.e. in its experience linked with this specific part of the universe. Therefore, it is difficult to understand fundamentally proverbs of a foreign language and to identify the relevant frames (sphere of use) of the respective general semantic domain. Its study implies (along with that of the language) the study of the ethnic “metaphorical thinking”, practically ethnopsychology of the people that created it.

The ERP correlates of self-knowledge in ageing

Self-knowledge is a type of personal semantic knowledge that concerns one’s self-image and personal identity. It has most often been operationalized as the summary of one’s personality traits (“I am a stubborn person”). Interestingly, recent studies have revealed that the neural correlates of self-knowledge can be dissociated from those of general semantic and episodic memory in young adults. However, studies of “dedifferentiation” or loss of distinctiveness of neural representations in ageing suggest that the neural correlates of self-knowledge might be less distinct from those of semantic and episodic memory in older adults. We investigated this question in an event-related potential (ERP) study with 28 young and 26 older adults while they categorised personality traits for their self-relevance (self-knowledge conditions), and their relevance to certain groups of people (general semantic condition). Participants then performed a recognition test for previously seen traits (episodic condition). The amplitude of the late positive component (LPC), associated with episodic recollection processes, differentiated the self-knowledge, general semantic, and episodic conditions in young adults, but not in older adults. However, in older adults, participants with higher composite episodic memory scores had more differentiated LPC amplitudes across experimental conditions. Moreover, consistent with the fact that age-related neural dedifferentiation may be material and region specific, in both age groups some differences between memory types were observed for the N400 component, associated with semantic processing. Taken together, these findings suggest that declarative memory subtypes are less distinct in ageing, but that the amount of differentiation varies with episodic memory function.

Modelling General Properties of Nouns by Selectively Averaging Contextualised Embeddings

While the success of pre-trained language models has largely eliminated the need for high-quality static word vectors in many NLP applications, static word vectors continue to play an important role in tasks where word meaning needs to be modelled in the absence of linguistic context. In this paper, we explore how the contextualised embeddings predicted by BERT can be used to produce high-quality word vectors for such domains, in particular related to knowledge base completion, where our focus is on capturing the semantic properties of nouns. We find that a simple strategy of averaging the contextualised embeddings of masked word mentions leads to vectors that outperform the static word vectors learned by BERT, as well as those from standard word embedding models, in property induction tasks. We notice in particular that masking target words is critical to achieve this strong performance, as the resulting vectors focus less on idiosyncratic properties and more on general semantic properties. Inspired by this view, we propose a filtering strategy which is aimed at removing the most idiosyncratic mention vectors, allowing us to obtain further performance gains in property induction.

Syntactic and Semantic Influences on the Time Course of Relative Clause Processing: The Role of Language Dominance

We conducted a visual world eye-tracking experiment with highly proficient Spanish-English bilingual adults to investigate the effects of relative language dominance, operationalized as a continuous, multidimensional variable, on the time course of relative clause processing in the first-learned language, Spanish. We found that participants exhibited two distinct processing preferences: a semantically driven preference to assign agency to referents of lexically animate noun phrases and a syntactically driven preference to interpret relative clauses as subject-extracted. Spanish dominance was found to exert a distinct influence on each of these preferences, gradiently attenuating the semantic preference while gradiently exaggerating the syntactic preference. While these results might be attributable to particular properties of Spanish and English, they also suggest a possible generalization that greater dominance in a language increases reliance on language-specific syntactic processing strategies while correspondingly decreasing reliance on more domain-general semantic processing strategies.