Semantic Framework Research Articles

Analysis of Dataset Limitations in Semantic Knowledge-Driven Multi-Variant Machine Translation

In this study, we explore the implications of dataset limitations in semantic knowledge-driven machine translation (MT) for intelligent virtual assistants (IVA). Our approach diverges from traditional single-best translation techniques, utilizing a multi-variant MT method that generates multiple valid translations per input sentence through a constrained beam search. This method extends beyond the typical constraints of specific verb ontologies, embedding within a broader semantic knowledge framework. We evaluate the performance of multi-variant MT models in translating training sets for Natural Language Understanding (NLU) models. These models are applied to semantically diverse datasets, including a detailed evaluation using the standard MultiATIS++ dataset. The results from this evaluation indicate that while multi-variant MT method is promising, its impact on improving intent classification (IC) accuracy is limited when applied to conventional datasets such as MultiATIS++. However, our findings underscore that the effectiveness of multi-variant translation is closely associated with the diversity and suitability of the datasets utilized. Finally, we provide an in-depth analysis focused on generating variant-aware NLU datasets. This analysis aims to offer guidance on enhancing NLU models through semantically rich and variant-sensitive datasets, maximizing the advantages of multi-variant MT.

A Cognitive Semantic Account of the Preposition “‘bra” in Modern Standard Arabic with Reference to the English “Through”

Prepositions are crucial grammatical devices for indicating location, movement, time, and other meanings. A case in point is the Arabic preposition “‘abra,” which expresses different spatial, temporal, and abstract meanings. Adopting a cognitive semantic perspective, this study investigates ‘abra, its diverse spatial meanings, and metaphorical extensions. It also uses cognitive constructs such as image schemas, prototypes, and conceptual metaphor to demonstrate how various senses of ‘abra can be described based on perceptual properties and cognitive characteristics. In addition, this study compares the meanings of ‘abra with those of its English equivalent (through). This research uses authentic examples from Modern Standard Arabic available through the Sketch Engine website and illustrations created by the researcher. The analytical description involves the basic schema, the prototypical meaning, and the metaphorical extensions. The findings reveal that ‘abra and “through” share a basic image schema and prototypical sense with slight differences resulting from varying conceptualizations. The metaphorical extensions are also similar. The distinctions between the two mostly pertain to the semantic scope in addition to the fact that ‘abra also covers the meaning of the preposition “across.” Generally, this analysis asserts that prepositional meanings are best described using a cognitive semantic framework. The results may be useful in the fields of lexicon, second learning and teaching, as well as translation.

Modeling High-Order Relationships: Brain-Inspired Hypergraph-Induced Multimodal-Multitask Framework for Semantic Comprehension.

Semantic comprehension aims to reasonably reproduce people's real intentions or thoughts, e.g., sentiment, humor, sarcasm, motivation, and offensiveness, from multiple modalities. It can be instantiated as a multimodal-oriented multitask classification issue and applied to scenarios, such as online public opinion supervision and political stance analysis. Previous methods generally employ multimodal learning alone to deal with varied modalities or solely exploit multitask learning to solve various tasks, a few to unify both into an integrated framework. Moreover, multimodal-multitask cooperative learning could inevitably encounter the challenges of modeling high-order relationships, i.e., intramodal, intermodal, and intertask relationships. Related research of brain sciences proves that the human brain possesses multimodal perception and multitask cognition for semantic comprehension via decomposing, associating, and synthesizing processes. Thus, establishing a brain-inspired semantic comprehension framework to bridge the gap between multimodal and multitask learning becomes the primary motivation of this work. Motivated by the superiority of the hypergraph in modeling high-order relations, in this article, we propose a hypergraph-induced multimodal-multitask (HIMM) network for semantic comprehension. HIMM incorporates monomodal, multimodal, and multitask hypergraph networks to, respectively, mimic the decomposing, associating, and synthesizing processes to tackle the intramodal, intermodal, and intertask relationships accordingly. Furthermore, temporal and spatial hypergraph constructions are designed to model the relationships in the modality with sequential and spatial structures, respectively. Also, we elaborate a hypergraph alternative updating algorithm to ensure that vertices aggregate to update hyperedges and hyperedges converge to update their connected vertices. Experiments on the dataset with two modalities and five tasks verify the effectiveness of HIMM on semantic comprehension.

Psycho-semantic analysis of job advertisements as a communicative tool of educational marketing

Introduction. The relevance of the study of job advertisements for teachers and psychologists is associated with the need to find tools for regulating communication within the modern educational environment with dynamic, constantly changing working conditions. The publication of employer requirements sets the direction for changing the criteria for the success of the specialist, which changes the entire educational environment. The purpose of the study: to identify the categorical semantic structure of job advertisements for teachers (using the example of the "Infrastructure of Scientific Research Data" platform). Materials and methods. The object of the study was a corpus of teacher job vacancies from the "Infrastructure of Scientific Research Data" platform, published in 2023 in 83 regions of Russia (5561 vacancies). The job postings included descriptions of candidate requirements, information about the offered salary, and the region in which the vacancy was available. As a comparison group, a corpus of psychologist job vacancies (2076 vacancies) was considered. The following methods were used in the study: 1) Latent Dirichlet Allocation (LDA) method for identifying and analyzing the structure of requirements presented in teacher and psychologist job vacancies; 2) Term Frequency-Inverse Document Frequency (TF-IDF) method for categorical semantic analysis of the structure of teacher and psychologist job vacancies and description of the semantic framework (frame) of the job description; 3) Descriptive statistical methods for analyzing the needs of professionals in helping professions in different regions of Russia and the conditions offered to them. The results of the study. The regions of the Russian Federation with the greatest need for teachers and psychologists have been identified. Based on the statistical significance in the collection of text documents, the most demanded skills of teachers by employers are identified: methodological development and design of an individual educational route (0.167), work with children with disabilities (0.144), organization of the educational process (0.089). For vacancies of psychologists, the most significant were: work on investigative actions (0.319), methodical work in an educational organization (0.187), psychological and pedagogical support for successful development and training (0.129). The main meanings governing professional pedagogical activity are more focused on meta-skills based on the values of inclusion and continuous improvement of education, as well as on flexible skills (Soft skills) that increase competitiveness in various fields of pedagogical activity. In the vacancies of psychologists, professional skills (Hard skills) are relevant, applicable in specific areas, which somewhat narrows the scope of professional implementation. Conclusion. The results of the empirical study allow for considering job vacancy advertisements as one of the communicative tools of educational marketing. Analyzing the significant professional skills from the employer's perspective will enable the development of recommendations for the development of professional communication for teachers, which can be considered in the system of continuous professional education, as well as in regulating interaction within the educational environment. Automated analysis of job vacancies can become a successful interactive diagnostic tool for educational marketing, necessary for the development of the educational space and the improvement of the competitiveness of educational institutions.

Deep-Pelagic Fishes Are Anything But Similar: A Global Synthesis.

Deep-pelagic fishes are among the most abundant vertebrates on Earth. They play a critical role in sequestering carbon, providing prey for harvestable fishing stocks and linking oceanic layers and trophic levels. However, knowledge of these fishes is scarce and fragmented, hampering the ability of both the scientific community and stakeholders to address them effectively. While modelling approaches incorporating these organisms have advanced, they often oversimplify their functional and ecological diversity, potentially leading to misconceptions. To address these gaps, this synthesis examines the biodiversity and ecology of global deep-pelagic fishes. We review pelagic ecosystem classifications and propose a new semantic framework for deep-pelagic fishes. We evaluate different sampling methods, detailing their strengths, limitations and complementarities. We provide an assessment of the world's deep-pelagic fishes comprising 1554 species, highlighting major groups and discussing regional variability. By describing their morphological, behavioural and ecological diversity, we show that these organisms are far from homogeneous. Building on this, we call for a more realistic approach to the ecology of deep-pelagic fishes transitioning between very different ecological niches during diel vertical migrations. To facilitate this, we introduce the concept of 'diel-modulated realised niche' and propose a conceptual model synthesising the multiple drivers responsible for such transitions.

Epistemic Logics for Relevant Reasoners

We present a neighbourhood-style semantic framework for modal epistemic logic modelling agents who process information using relevant logic. The distinguishing feature of the framework in comparison to relevant modal logic is that the environment the agent is situated in is assumed to be a classical possible world. This framework generates two-layered logics combining classical logic on the propositional level with relevant logic in the scope of modal operators. Our main technical result is a general soundness and completeness theorem.

LESA-Net: Semantic segmentation of multi-type road point clouds in complex agroforestry environment

Point-cloud semantic segmentation is a visual task essential for agricultural robots to comprehend natural agroforestry environments. However, owing to the extremely large amount of point-cloud data in agroforestry environments, learning effective features for semantic segmentation from large-scale point clouds is challenging. Therefore, to address this issue and achieve accurate semantic segmentation of different types of road-surface point clouds in large-scale agroforestry environments, this study proposes a point-cloud semantic segmentation network framework based on double-distance self-attention. First, a point-cloud local feature enhancement module is proposed. This module primarily extends the receptive field and enhances the generalizability of multidimensional features by incorporating reflection intensity information and a spatial feature-encoding block that is enhanced with contextual semantic information. Second, we introduce a dual-distance attention pooling (DDAPS) block based on the self-attention mechanism. This block initially learns the feature representation of the local neighborhood of each point through the self-attention mechanism. Then, it uses the DDAPS block to aggregate more discriminative local neighborhood point features. Finally, extensive experimental results on large-scale point-cloud datasets, SemanticKITTI and RELLIS-3D, demonstrate that our algorithm outperforms similar algorithms in large-scale agroforestry environments.

Building a Turkish UCCA dataset

Abstract Semantic representation is the task of conveying the meaning of a natural language utterance by converting it to a logical form that can be processed and understood by machines. It is utilised by many applications in natural language processing (NLP), particularly in tasks relevant to natural language understanding (NLU). Due to the widespread use of semantic parsing in NLP, many semantic representation schemes with different forms have been proposed; Universal Conceptual Cognitive Annotation (UCCA) is one of them. UCCA is a cross-lingual semantic annotation framework that allows easy annotation without requiring substantial linguistic knowledge. UCCA-annotated datasets have been released so far for English, French, German, Russian, and Hebrew. In this paper, we present a UCCA-annotated Turkish dataset of 400 sentences that are obtained from the METU-Sabanci Turkish Treebank. We provide the UCCA annotation specifications defined for the Turkish language so that it can be extended further. We followed a semi-automatic annotation approach, where an external semantic parser is utilised for the initial annotation of the dataset, which is manually revised by two annotators. We used the same semantic parser model to evaluate the dataset with zero-shot and few-shot learning, demonstrating that even a small sample set from the target language in the training data has a notable impact on the performance of the parser (15.6% and 2.5% gain over zero-shot for labelled and unlabelled results, respectively).

A domain adaptation framework for cross-modality SAR 3D reconstruction point clouds segmentation utilizing LiDAR data

Synthetic aperture radar (SAR) 3D point clouds reconstruction can eliminate the problems of layover in 2D SAR image projections, the recognition of the reconstructed point cloud can significantly enhance target identification and information extraction. Current SAR 3D point cloud segmentation methods, based on traditional machine learning clustering approaches, suffer from poor automation and accuracy. However, the absence of publicly labeled SAR datasets impedes the progress of deep learning-based SAR 3D point cloud segmentation methods. To tackle the aforementioned challenges, we introduce, for the first time, an alternative training approach for SAR 3D point cloud segmentation using LiDAR annotated data, which offers a more abundant sample pool, thus alleviating the lack of training sets. Nevertheless, a segmentation model trained on LiDAR point clouds exhibits a significant decline in performance when directly applied to SAR 3D reconstructed point clouds due to the cross-domain discrepancy. This research presents a pioneering domain adaptation 3D semantic segmentation framework to implement cross-modal learning for SAR point clouds. In our scheme, a simple yet effective technique is developed to achieve segmentation of SAR double-bounce scattering regions called SARDBS-Mix, which employs a mixing strategy, to capture the distinctive reflection characteristics of SAR data. Furthermore, we implement approaches including center alignment and normalization, local augmentation, and weighted cross entropy to mitigate LiDAR and SAR domain gap and class imbalances. The experimental results validate the feasibility and effectiveness of the proposed method for SAR 3D reconstructed point cloud segmentation.

Key Research Areas for Building and Deploying a Common Data Model for an Intensive Medicine Data Space in Europe and Beyond.

Key Research Areas (KRAs) were identified to establish a semantic interoperability framework for intensive medicine data in Europe. These include assessing common data model value, ensuring smooth data interoperability, supporting data standardization for efficient dataset use, and defining anonymization requirements to balance data protection and innovation.

Automated Verification of Higher-Order Probabilistic Programs via a Dependent Refinement Type System

Verification of higher-order probabilistic programs is a challenging problem. We present a verification method that supports several quantitative properties of higher-order probabilistic programs. Usually, extending verification methods to handle the quantitative aspects of probabilistic programs often entails extensive modifications to existing tools, reducing compatibility with advanced techniques developed for qualitative verification. In contrast, our approach necessitates only small amounts of modification, facilitating the reuse of existing techniques and implementations. On the theoretical side, we propose a dependent refinement type system for a generalised higher-order fixed point logic (HFL). Combined with continuation-passing style encodings of properties into HFL, our dependent refinement type system enables reasoning about several quantitative properties, including weakest pre-expectations, expected costs, moments of cost, and conditional weakest pre-expectations for higher-order probabilistic programs with continuous distributions and conditioning. The soundness of our approach is proved in a general setting using a framework of categorical semantics so that we don’t have to repeat similar proofs for each individual problem. On the empirical side, we implement a type checker for our dependent refinement type system that reduces the problem of type checking to constraint solving. We introduce admissible predicate variables and integrable predicate variables to constrained Horn clauses (CHC) so that we can soundly reason about the least fixed points and samplings from probability distributions. Our implementation demonstrates that existing CHC solvers developed for non-probabilistic programs can be extended to a solver for the extended CHC with only small efforts. We also demonstrate the ability of our type checker to verify various concrete examples.

A Global Features and Local Features-Jointly Fused Deep Semantic Learning Framework for Error Detection of Machine Translation

A Global Features and Local Features-Jointly Fused Deep Semantic Learning Framework for Error Detection of Machine Translation

A Novel Global-Local Feature Aggregation Framework for Semantic Segmentation of Large-Format High-Resolution Remote Sensing Images

In high-resolution remote sensing images, there are areas with weak textures such as large building roofs, which occupy a large number of pixels in the image. These areas pose a challenge for traditional semantic segmentation networks to obtain ideal results. Common strategies like downsampling, patch cropping, and cascade models often sacrifice fine details or global context, resulting in limited accuracy. To address these issues, a novel semantic segmentation framework has been designed specifically for large-format high-resolution remote sensing images by aggregating global and local features in this paper. The framework consists of two branches: one branch deals with low-resolution downsampled images to capture global features, while the other branch focuses on cropped patches to extract high-resolution local details. Also, this paper introduces a feature aggregation module based on the Transformer structure, which effectively aggregates global and local information. Additionally, to save GPU memory usage, a novel three-step training method has been developed. Extensive experiments on two public datasets demonstrate the effectiveness of the proposed approach, with an IoU of 90.83% on the AIDS dataset and 90.30% on the WBDS dataset, surpassing state-of-the-art methods such as DANet, DeepLab v3+, U-Net, ViT, TransUNet, CMTFNet, and UANet.

Linguistic-Driven Partial Semantic Relevance Learning for Skeleton-Based Action Recognition.

Skeleton-based action recognition, renowned for its computational efficiency and indifference to lighting variations, has become a focal point in the realm of motion analysis. However, most current methods typically only extract global skeleton features, overlooking the potential semantic relationships among various partial limb motions. For instance, the subtle differences between actions such as "brush teeth" and "brush hair" are mainly distinguished by specific elements. Although combining limb movements provides a more holistic representation of an action, relying solely on skeleton points proves inadequate for capturing these nuances. Therefore, integrating detailed linguistic descriptions into the learning process of skeleton features is essential. This motivates us to explore integrating fine-grained language descriptions into the learning process of skeleton features to capture more discriminative skeleton behavior representations. To this end, we introduce a new Linguistic-Driven Partial Semantic Relevance Learning framework (LPSR) in this work. While using state-of-the-art large language models to generate linguistic descriptions of local limb motions and further constrain the learning of local motions, we also aggregate global skeleton point representations and textual representations (which generated from an LLM) to obtain a more generalized cross-modal behavioral representation. On this basis, we propose a cyclic attentional interaction module to model the implicit correlations between partial limb motions. Numerous ablation experiments demonstrate the effectiveness of the method proposed in this paper, and our method also obtains state-of-the-art results.

Coalgebraic Satisfiability Checking for Arithmetic $\mu$-Calculi

The coalgebraic $\mu$-calculus provides a generic semantic framework for fixpoint logics over systems whose branching type goes beyond the standard relational setup, e.g. probabilistic, weighted, or game-based. Previous work on the coalgebraic $\mu$-calculus includes an exponential-time upper bound on satisfiability checking, which however relies on the availability of tableau rules for the next-step modalities that are sufficiently well-behaved in a formally defined sense; in particular, rule matches need to be representable by polynomial-sized codes, and the sequent duals of the rules need to absorb cut. While such rule sets have been identified for some important cases, they are not known to exist in all cases of interest, in particular ones involving either integer weights as in the graded $\mu$-calculus, or real-valued weights in combination with non-linear arithmetic. In the present work, we prove the same upper complexity bound under more general assumptions, specifically regarding the complexity of the (much simpler) satisfiability problem for the underlying one-step logic, roughly described as the nesting-free next-step fragment of the logic. The bound is realized by a generic global caching algorithm that supports on-the-fly satisfiability checking. Notably, our approach directly accommodates unguarded formulae, and thus avoids use of the guardedness transformation. Example applications include new exponential-time upper bounds for satisfiability checking in an extension of the graded $\mu$-calculus with polynomial inequalities (including positive Presburger arithmetic), as well as an extension of the (two-valued) probabilistic $\mu$-calculus with polynomial inequalities.

Orientational Metaphor in Central Kurdish

This study is entitled "Orientational Metaphor in Central Kurdish". It attempts to explicate how non-spatial concepts are metaphorised in terms of various aspects of space in Central Kurdish. The main aim of the study is to interpret spatial metaphorical expressions that express the cognitive processes of meaning-making. The study adopts the Conceptual Metaphor Theory (CMT)within the framework of cognitive semantics. It deals with orientational metaphor as a specific type of conceptual metaphor. In cognitive semantics, language is a mirror of thought; linguistic expressions are manifestations of conceptual mappings structured in the mind. The data of the study is taken from ordinary everyday Kurdish. The data analysis illuminates the spatialization of a wide variety of abstract concepts through metaphorical uses of space. The study is in line with the cognitive linguists in that linguistic expressions that represent concepts are not semantically independent, as the Anglo-American philosophical tradition claimed, but are structured with respect to one another. Spatial concepts are first structured in thought in the form of a formula based on physical and socio-cultural experiences, and then expressed via linguistic metaphors; accordingly, metaphor is a cognitive phenomenon, it is in thought, not in language. It is patterns of thought that are echoed in patterns of language.

A computable and compositional semantics for hybrid systems

Hybrid Systems are systems having a mixed discrete and continuous behaviour that cannot be characterised faithfully using either discrete or continuous models only.Due to the intrinsic complexity of hybrid systems, it is highly desirable to describe them compositionally, where large systems are seen as the composition of several simpler subparts that are studied independently. Furthermore, since several problems, including reachability, are undecidable for hybrid systems, the availability of approximate decision procedures is another important feature.In this paper, we propose a purely behavioural formalism for hybrid systems that is compositional and supports approximate decision procedures. The formalism is abstract and sufficiently general to subsume or be a semantic framework for most of the concrete formalisms or languages proposed in the literature.

A unified deep semantic expansion framework for domain-generalized person re-identification

Supervised Person Re-identification (Person ReID) methods have achieved excellent performance when training and testing within one camera network. However, they usually suffer from considerable performance degradation when applied to different camera systems. In recent years, many Domain Adaptation Person ReID methods have been proposed, achieving impressive performance without requiring labeled data from the target domain. However, these approaches still need the unlabeled data of the target domain during the training process, making them impractical in many real-world scenarios. Our work focuses on the more practical Domain Generalized Person Re-identification (DG-ReID) problem. Given one or more source domains, it aims to learn a generalized model that can be applied to unseen target domains. One promising research direction in DG-ReID is the use of implicit deep semantic feature expansion, and our previous method, Domain Embedding Expansion (DEX), is one such example that achieves powerful results in DG-ReID. However, in this work we show that DEX and other similar implicit deep semantic feature expansion methods, due to limitations in their proposed loss function, fail to reach their full potential on large evaluation benchmarks as they have a tendency to saturate too early. Leveraging on this analysis, we propose Unified Deep Semantic Expansion, our novel framework that unifies implicit and explicit semantic feature expansion techniques in a single framework to mitigate this early over-fitting and achieve a new state-of-the-art (SOTA) in all DG-ReID benchmarks. Further, we apply our method on more general image retrieval tasks, also surpassing the current SOTA in all of these benchmarks by wide margins.

Semantic model-based large-scale deployment of AI-driven building management applications

Digitalization and Artificial Intelligent (AI) are revolutionizing building operation management. The abundance of data generated with the digitalization of buildings in the whole lifecycle can be harnessed to enhance building operational efficiency through data-driven control and optimization applications. However, the heterogeneity of data across building datasets hampers data interactivity and interoperability, presenting obstacles to the large-scale deployment of AI-enabled data-driven solutions. A semantic model-based framework is developed to integrate multi-sources data from buildings' air-conditioning system, supporting the large-scale deployment of AI-enabled data-driven building management applications. Both static and temporal data from multi sources are stored in the database guided by the semantic model. To demonstrate the framework's effectiveness, a building cooling load prediction application is implemented and evaluated across three typical buildings. The successful deployment of the proposed semantic model-based framework demonstrates its potential for facilitating large-scale deployment of AI-enabled data-driven applications in building sector.

SEFANet: Semantic enhanced with feature alignment network for semantic segmentation

To address the issues of poor segmentation accuracy and insensitivity to details in semantic segmentation, this paper proposes a novel image semantic segmentation framework SEFANet. Specifically, SEFANet adopts encoder-decoder structure, and incorporates a novel perceptual enhancement mechanism called Multi-scale Spatial Integration Module (MSIM) at the encoder. MSIM is based on group convolution to boost spatial semantic features and refine spatial-gradient semantic features within the multi-scale structure. This module enhances feature extraction across different network levels, leading to improved edge detection and segmentation abilities. In the decoder, SEFANet introduces a pixel-level Interleaved Feature Alignment Module (IFAM), which leverages rich semantic information in low-dimensional features and the strategy of Semantic Offset Field. Meanwhile, IFAM warps the high-dimensional feature map into low-dimensional features, completing the calibration process through convolution operations. Experimental results on the Pascal VOC2012 val dataset and the Cityscapes val dataset confirm the effectiveness and generalization of the proposed semantic segmentation. Additionally, the results further demonstrate that SEFANet improves the poor segmentation accuracy and insensitivity to details, and achieves a competitive performance compared with other semantic segmentation methods.