Contextual Features Research Articles

Pre-syntactic impoverishment: Premature exponence in Harmonic Serialism

Impoverishment is a feature deletion operation that is standardly taken to require a post-syntactic approach to inflectional morphology (like, in particular, Distributed Morphology, as in Halle & Marantz 1993 ). Given that by now a lot of empirical evidence in support of such an operation has been accumulated, impoverishment can be viewed as a good argument in favour of the post-syntactic approach, given a Chomskyan Y-model of grammar in which inflectional morphology is considered an autonomous component of grammar (i.e., not subsumed under syntax proper). Against this background, the main goal of the present contribution is to show that a pre-syntactic approach to impoverishment becomes available if the complete well-formed sets of morpho-syntactic features that are required in any non-syntactic realizational approach to inflectional morphology (cf. Stump 2001 ; Inkelas 2016 ) are not given a declarative characterization (as is normally assumed), but arise incrementally, by iterated addition of features. On this view, pre-syntactic impoverishment is actually premature exponence brought about by late addition of context features, and the interaction between impoverishment and morphological exponence is not one of bleeding (as in post-syntactic approaches) but one of counter-feeding. The new concept of impoverishment is implemented in a pre-syntactic model of inflectional morphology developed in Müller (2020) , which is based on Harmonic Serialism, a derivational version of Optimality Theory (cf. Prince & Smolensky 2004 ). Finally, while pre-syntactic impoverishment is basically a mirror image of post-syntactic impoverishment, it turns out to be more restrictive since it ceteris paribus predicts that there is no impoverishment of inherent features: Features that are not added cannot be added late. I suggest that this prediction is confirmed by evidence from gender features and inflection class features.

Ethnic studies and student development: Cultivating racially marginalized adolescents' critical consciousness.

There is debate around offering ethnic studies to high school students. Ethnic studies connects learning to students' lives and analyzes the workings of racism to construct avenues toward equity. As the debate unfolds, it is critical to examine ethnic studies' implications for youth development and the mechanisms that link it to student outcomes. One of ethnic studies' long-stated goals is fostering students' critical consciousness. Critical consciousness refers to critical reasoning around inequality (critical reflection), motivation to challenge inequality (critical motivation), and action taken to disrupt inequality (critical action). Little research has examined youth critical consciousness development within ethnic studies-a consciousness-raising system. Consequently, this longitudinal mixed-methods study examines students' critical consciousness development in ethnic studies and sheds light on the contextual characteristics (i.e., critical school socialization) that foster critical consciousness. Analyses of 459 ninth-grade students' (52% girls, 4% nonbinary; 1% Asian, 1% Black, 4% multiracial, 64% Latinx, 7% Native American, 15% described their own race, 7% skipped the question; Mage = 13.92) survey data, and focus group data with 19 students, revealed that ethnic studies-enrolled students grew more in their critical reflection than nonenrolled students. However, the quantitative data demonstrated decreasing critical motivation among all students, whereas the qualitative data suggested emergent critical motivation among ethnic studies-enrolled students. Furthermore, critical school socialization and teacher pedagogy were key to ethnic studies consciousness-raising. Altogether, this study highlights that ethnic studies fosters youth critical consciousness-a worthwhile outcome that should be considered in policy debates about ethnic studies. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

HIF1α-regulated glycolysis promotes activation-induced cell death and IFN-γ induction in hypoxic T cells

Hypoxia is a common feature in various pathophysiological contexts, including tumor microenvironment, and IFN-γ is instrumental for anti-tumor immunity. HIF1α has long been known as a primary regulator of cellular adaptive responses to hypoxia, but its role in IFN-γ induction in hypoxic T cells is unknown. Here, we show that the HIF1α-glycolysis axis controls IFN-γ induction in both human and mouse T cells, activated under hypoxia. Specific deletion of HIF1α in T cells (Hif1α–/–) and glycolytic inhibition suppresses IFN-γ induction. Conversely, HIF1α stabilization by hypoxia and VHL deletion in T cells (Vhl–/–) increases IFN-γ production. Hypoxic Hif1α–/– T cells are less able to kill tumor cells in vitro, and tumor-bearing Hif1α–/– mice are not responsive to immune checkpoint blockade (ICB) therapy in vivo. Mechanistically, loss of HIF1α greatly diminishes glycolytic activity in hypoxic T cells, resulting in depleted intracellular acetyl-CoA and attenuated activation-induced cell death (AICD). Restoration of intracellular acetyl-CoA by acetate supplementation re-engages AICD, rescuing IFN-γ production in hypoxic Hif1α–/– T cells and re-sensitizing Hif1α–/– tumor-bearing mice to ICB. In summary, we identify HIF1α-regulated glycolysis as a key metabolic control of IFN-γ production in hypoxic T cells and ICB response.

A hybrid BiLSTM-CNN deep learning model for Chinese sentiment analysis of online car reviews

Although sentiment analysis with datasets in English has achieved significant progress, there are still relatively few studies on sentiment analysis in the area of Chinesecar review texts. In addition, the existing Chinese text sentiment analysis methods cannot simultaneously extract the contextual features and the local semantic information from the review texts. To address the above issues, a hybrid deep learning model namely BiLSTM-CNN was designed and validated with Chinese car review texts. We trained word vectors with the CBOW model by combining the Chinese Wikipedia corpus with other open-source car-related corpora. Such word vectors include car-specific vocabulary, which improves sentimental classification accuracy. Experimental results show that the performance indices of the deep learning models (CNN, LSTM, BiLSTM) are much better than the KNN, SVM, Naïve Bayes, and RF model, which is attributed to deep learning's powerful feature extraction ability and nonlinear fitting ability. Furthermore, when compared to deep learning models such as CNN, LSTM, BiLSTM, and LSTM-CNN, the suggested hybrid BiLSTM-CNN model outperformed the others. The results may provide references for consumers to buy a car and for car companies to optimize their products.

Review-Based Explainable Recommendations: A Transparency Perspective

With the prevalence of deep learning based embedding approaches, recommender systems have become a proven and indispensable tool in various information filtering applications. However, many of them remain difficult to diagnose what aspects of the deep models’ input drive the final ranking decision, thus, they cannot often be understood by human stakeholders. In this paper, we investigate the dilemma between recommendation and explainability, and show that by utilizing the contextual features (e.g., item reviews from users), we can design a series of explainable recommender systems without sacrificing their performance. In particular, we propose three types of explainable recommendation strategies with gradual changes in model transparency: whitebox, graybox, and blackbox. Each strategy explains its ranking decisions via different mechanisms: attention weights, adversarial perturbations, and counterfactual perturbations. We apply these explainable models on five real-world data sets under the contextualized setting where the users provided explicit reviews of the items. The empirical results show that our models can achieve highly competitive ranking performance, and meanwhile, generate accurate and effective explanations in terms of numerous quantitative metrics and qualitative visualizations.

Drug–drug interaction extraction based on multimodal feature fusion by Transformer and BiGRU

Understanding drug–drug interactions (DDIs) plays a vital role in the fields of drug disease treatment, drug development, preventing medical error, and controlling health care-costs. Extracting potential from biomedical corpora is a major complement of existing DDIs. Most existing DDI extraction (DDIE) methods do not consider the graph and structure of drug molecules, which can improve the performance of DDIE. Considering the different advantages of bi-directional gated recurrent units (BiGRU), Transformer, and attention mechanisms in DDIE tasks, a multimodal feature fusion model combining BiGRU and Transformer (BiGGT) is here constructed for DDIE. In BiGGT, the vector embeddings of medical corpora, drug molecule topology graphs, and structure are conducted by Word2vec, Mol2vec, and GCN, respectively. BiGRU and multi-head self-attention (MHSA) are integrated into Transformer to extract the local–global contextual DDIE features, which is important for DDIE. The extensive experiment results on the DDIExtraction 2013 shared task dataset show that the BiGGT-based DDIE method outperforms state-of-the-art DDIE approaches with a precision of 78.22%. BiGGT expands the application of multimodal deep learning in the field of multimodal DDIE.

Innovations in abortion care in France, Germany and the UK following the COVID-19 pandemic

Abstract Background Through an international comparison of health systems and policies, we aim at assessing the contextual and circumstantial characteristics of abortion care innovation in three European countries following the upheaval of the COVID-19 pandemic. The three contrasting care models pre-pandemic constitute an interesting baseline for observing different systems’ reaction to health crises. Methods We conducted a review of the recent (2020 onward) peer-reviewed and grey literature, including national policy documents and guidelines. We then carried a comparative analysis of innovations implemented during the pandemic and their sustainability post-crisis. Results In France and the UK, where abortion care was more accessible before the pandemic than in Germany, a series of innovations were officially introduced to facilitate access during the pandemic. They included teleconsultations and the mailing of abortion medication. Most changes have been sustained since then, contributing to addressing historic and systemic health inequities in terms of access (e.g. due to geography and the existence of “medical deserts”). In Germany, innovations during the pandemic have been crafted by the civil society through NGOs offering for the first time teleabortion services (e.g. Balance Berlin and Doctors for Choice). The legitimation of those practices by professional bodies is still in the making, highlighting the limitations of the catalyst role of crises in the face of cultural barriers and resistance of health systems to innovation. Conclusions The COVID-19 pandemic provoked or accelerated innovation in terms of abortion care in France, Germany and the UK. Yet, its sustainability, as well as its scaling up, remains fragile, especially in Germany, where the disruptive approach of civil society organisations has not yet found its way into mainstream health care services. Key messages • International comparison of health systems’ reaction to the COVID-19 pandemic in terms of abortion care shows different paths to innovation. • The sustainability of innovation beyond the first pandemic years pertains to the wider debate on equity in access to abortion care.

Medical image segmentation with UNet-based multi-scale context fusion

Histopathological examination holds a crucial role in cancer grading and serves as a significant reference for devising individualized patient treatment plans in clinical practice. Nevertheless, the distinctive features of numerous histopathological image targets frequently contribute to suboptimal segmentation performance. In this paper, we propose a UNet-based multi-scale context fusion algorithm for medical image segmentation, which extracts rich contextual information by extracting semantic information at different encoding stages and assigns different weights to the semantic information at different scales through TBSFF module to improve the learning ability of the network for features. Through multi-scale context fusion and feature selection networks, richer semantic features and detailed information are extracted. The target can be more accurately segmented without significantly increasing the extra overhead. The results demonstrate that our algorithm achieves superior Dice and IoU scores with a relatively small parameter count. Specifically, on the GlaS dataset, the Dice score is 90.56, and IoU is 83.47. For the MoNuSeg dataset, the Dice score is 79.07, and IoU is 65.98.

Service similarity measurement integrating Bi-LSTM contextual representation and attention mechanism for web service discovery

With the rapid growth and wide adoption of web services, service discovery is becoming commonly used to locate the optimal services that can meet the requirements of users. This paper proposes a bidirectional long short-term memory (Bi-LSTM) service discovery method that integrates contextual representation and attention mechanisms for web service discovery. To conduct the study, FastText word embedding with the N-gram feature is primarily used as the input of the Bi-LSTM model to obtain the context feature vector of the service function description. Moreover, the attention mechanism is employed to calculate the similarity of the service function description to gain the semantic feature vector of the correlation degree between services. Then, the multilayer perceptron network is used to mine the mapping relationship between semantic feature vectors and matching levels among services. Finally, for a given query, target services are retrieved by ranking candidate services according to the prediction cores through the matching method. The proposed method is examined against multiple evaluation metrics, including accuracy and error, and compared with the state-of-the-art searching approaches. The experimental results show that the proposed method is more effective and attains better relevant web services, which helps the request service find more accurate results. Furthermore, this study provides new insights for researchers into web service discovery.

Circumstantial Factors Among Kentucky Nurse Suicide Decedents, 2005 to 2019.

Background: Nurse suicide, a complex occupational health concern, is urgently in need of research due to the personal and occupational suicide risk factors experienced by nurses, namely mental health problems (e.g., depression), job problems, and substance misuse. Therefore, the study aims were to determine the contextual characteristics and circumstantial factors associated with nurse suicide in Kentucky. Methods: Secondary suicide data were obtained from the Kentucky Violent Death Reporting System (KYVDRS) from 2005 to 2019. Nurse suicide cases were identified using the Bureau of Labor Statistics Standard Occupational Classification. A mixed-methods analysis using descriptive statistics and qualitative evaluation was conducted to determine the distributions of demographic, injury, and weapon characteristics, followed by a qualitative analysis of the KYVDRS incident narrative text of nurse suicide decedents. Results: There were 88 decedents identified with a nursing occupation. The predominant means of death for male (59%) and female (45%) nurses involved firearms. Thematic analysis showed nurse decedents experienced a mental health problem (51%) and premeditated (50%) suicide preceding death. Incidentally, 51% of all cases experienced multiple circumstantial factors (e.g., relationship problems and premeditation) preceding their death by suicide. Discussion/Application to Practice: Nurse suicide decedents experienced multifactorial risk factors preceding their death by suicide, namely depression, premeditated suicide, and relationship problems. Occupational health practitioners and health care organizations can implement individual and organizational prevention efforts to help prevent nurse suicide.

Recurrence prediction of gastric cancer based on multi-resolution feature fusion and context information

Pathological images of gastric cancer serve as the gold standard for diagnosing this malignancy. However, the recurrence prediction task often encounters challenges such as insignificant morphological features of the lesions, insufficient fusion of multi-resolution features, and inability to leverage contextual information effectively. To address these issues, a three-stage recurrence prediction method based on pathological images of gastric cancer is proposed. In the first stage, the self-supervised learning framework SimCLR was adopted to train low-resolution patch images, aiming to diminish the interdependence among diverse tissue images and yield decoupled enhanced features. In the second stage, the obtained low-resolution enhanced features were fused with the corresponding high-resolution unenhanced features to achieve feature complementation across multiple resolutions. In the third stage, to address the position encoding difficulty caused by the large difference in the number of patch images, we performed position encoding based on multi-scale local neighborhoods and employed self-attention mechanism to obtain features with contextual information. The resulting contextual features were further combined with the local features extracted by the convolutional neural network. The evaluation results on clinically collected data showed that, compared with the best performance of traditional methods, the proposed network provided the best accuracy and area under curve (AUC), which were improved by 7.63% and 4.51%, respectively. These results have effectively validated the usefulness of this method in predicting gastric cancer recurrence.

Density‐Aware Diffusion Model for Efficient Image Dehazing

AbstractExisting image dehazing methods have made remarkable progress. However, they generally perform poorly on images with dense haze, and often suffer from unsatisfactory results with detail degradation or color distortion. In this paper, we propose a density‐aware diffusion model (DADM) for image dehazing. Guided by the haze density, our DADM can handle images with dense haze and complex environments. Specifically, we introduce a density‐aware dehazing network (DADNet) in the reverse diffusion process, which can help DADM gradually recover a clear haze‐free image from a haze image. To improve the performance of the network, we design a cross‐feature density extraction module (CDEModule) to extract the haze density for the image and a density‐guided feature fusion block (DFFBlock) to learn the effective contextual features. Furthermore, we introduce an indirect sampling strategy in the test sampling process, which not only suppresses the accumulation of errors but also ensures the stability of the results. Extensive experiments on popular benchmarks validate the superior performance of the proposed method. The code is released in https://github.com/benchacha/DADM.

Investigating the implementation of a novel approach to alcohol screening and brief intervention in Mexico: a mixed-methods study using pseudo-patients.

Low- and middle-income countries bear disproportionate burdens from excessive alcohol consumption, yet have fewer resources to identify and intervene with risky drinkers. Low-cost screening and brief intervention (SBI) models offer a tool for addressing this health problem and reducing disparities. In this mixed-methods study, trained pseudo-patients visited health clinics in Zacatecas, Mexico, where a novel SBI model was used with trained nonmedical health educators (HEs) conducting SBI in waiting areas. Pseudo-patients, who provided responses to the AUDIT-C screening items designed to trigger a brief intervention (BI), waited for HEs to engage them in an SBI encounter. Data on HEs' behaviors, SBI components provided, and contextual characteristics were coded from audio recordings of the encounters using an SBI checklist and from pseudo-patient interviews. Quantitative analyses examined the consistency in pseudo-patients' targeted AUDIT-C scores and those documented by HEs as well as the frequency of delivery of SBI components. Across 71 interactions, kappas between HEs' scores and the targeted AUDIT-C scores ranged from 0.33 to 0.45 across AUDIT-C items; it was 0.16 for the total AUDIT-C. In 41% of interactions, the HEs recorded total AUDIT-C scores that accurately reflected the targeted scores, 45% were below, and 14% exceeded them. Analyses of checklist items and transcripts showed that HEs demonstrated desired interpersonal skills (attentive, empathetic, professional) and provided general information regarding risks and recommendations about reducing consumption. In contrast, personalized BI components (exploring pseudo-patients' personal challenges and concerns about reducing drinking; making a plan) occurred much less frequently. Pseudo-patient interviews revealed contextual factors (noise, lack of privacy) that may have negatively affected SBI interactions. Using trained nonmedical persons to administer SBI holds promise to increase its reach. However, ongoing training and monitoring, prioritizing comprehensive BIs, eliminating contextual barriers, and electronic delivery of screening may help ensure high quality delivery.

Gendered opportunities across modes of political participation: a macro–micro analysis of the gender gap

We examine the role of ‘gendered opportunities’ for political participation by analysing original survey data covering nine European countries alongside relevant macro-level factors. We hypothesize that gender interacts differently with certain features of the broader context – here, government spending on public services – across various modes of participation, leading to ‘gendered opportunities’ for political engagement. By analysing data for three different modes of participation, namely voting, protest and consumerist participation, we show that the political context provides ‘gendered opportunities’ for political participation and that the gender gap is not homogeneously distributed across modes but varies depending on the repertoire. Our findings clearly highlight how the relationship between gender and the broader context for political participation is a complex one, suggesting that further work must consider the macro–micro linkages leading to differential gender inequalities in political participation.

Optimization Study of Coal Gangue Detection in Intelligent Coal Selection Systems Based on the Improved Yolov8n Model

To address the low recognition accuracy of models for coal gangue images in intelligent coal preparation systems—especially in identifying small target coal gangue due to factors such as camera angle changes, low illumination, and motion blur—we propose an improved coal gangue separation model, Yolov8n-improvedGD(GD—Gangue Detection), based on Yolov8n. The optimization strategy includes integrating the GCBlock(Global Context Block) from GCNet(Global Context Network) into the backbone network to enhance the model’s ability to capture long-range dependencies in images and improve recognition performance. The CGFPN (Contextual Guidance Feature Pyramid Network) module is designed to optimize the feature fusion strategy and enhance the model’s feature expression capabilities. The GSConv-SlimNeck architecture is employed to optimize computational efficiency and enhance feature map fusion capabilities, thereby improving the model’s robustness. A 160 × 160 scale detection head is incorporated to enhance the sensitivity and accuracy of small coal and gangue detection, mitigate the effects of low-quality data, and improve target localization accuracy.

HRTBDA: a network for post-disaster building damage assessment based on remote sensing images

ABSTRACT Efficient building damage assessment after disasters is vital for emergency response and loss evaluation, but the task is complicated by diverse building structures and complex environments. Traditional methods using Convolutional Neural Networks (CNNs) struggle to capture global contextual features, limiting damage categorization accuracy. To address this, we introduce the High-Resolution Transformer Architecture for Building Damage Assessment (HRTBDA), which enhances multi-scale feature extraction. A Cross-Attention-Based Spatial Fusion (CSF) module is proposed to utilize the attention mechanism, improving the model’s ability to identify detailed associations in damaged buildings. Additionally, we propose a deep convolution network matching optimization strategy that integrates a multilayer perceptron and expands the receptive field, enhancing global feature perception. HRTBDA’s performance was evaluated on two public datasets and compared with five recent frameworks. The model achieved an F1-score of 86.0% in building localization and 78.4% in damage assessment, with a 4.8% improvement in detecting minor damages. These results demonstrate HRTBDA’s potential for improving building damage assessment and highlight its significant advancements over existing methods.

THÖR-MAGNI: A large-scale indoor motion capture recording of human movement and robot interaction

We present a new large dataset of indoor human and robot navigation and interaction, called THÖR-MAGNI, that is designed to facilitate research on social human navigation: for example, modeling and predicting human motion, analyzing goal-oriented interactions between humans and robots, and investigating visual attention in a social interaction context. THÖR-MAGNI was created to fill a gap in available datasets for human motion analysis and HRI. This gap is characterized by a lack of comprehensive inclusion of exogenous factors and essential target agent cues, which hinders the development of robust models capable of capturing the relationship between contextual cues and human behavior in different scenarios. Unlike existing datasets, THÖR-MAGNI includes a broader set of contextual features and offers multiple scenario variations to facilitate factor isolation. The dataset includes many social human–human and human–robot interaction scenarios, rich context annotations, and multi-modal data, such as walking trajectories, gaze-tracking data, and lidar and camera streams recorded from a mobile robot. We also provide a set of tools for visualization and processing of the recorded data. THÖR-MAGNI is, to the best of our knowledge, unique in the amount and diversity of sensor data collected in a contextualized and socially dynamic environment, capturing natural human–robot interactions.

Cascade contour-enhanced panoptic segmentation for robotic vision perception.

Panoptic segmentation plays a crucial role in enabling robots to comprehend their surroundings, providing fine-grained scene understanding information for robots' intelligent tasks. Although existing methods have made some progress, they are prone to fail in areas with weak textures, small objects, etc. Inspired by biological vision research, we propose a cascaded contour-enhanced panoptic segmentation network called CCPSNet, attempting to enhance the discriminability of instances through structural knowledge. To acquire the scene structure, a cascade contour detection stream is designed, which extracts comprehensive scene contours using channel regulation structural perception module and coarse-to-fine cascade strategy. Furthermore, the contour-guided multi-scale feature enhancement stream is developed to boost the discrimination ability for small objects and weak textures. The stream integrates contour information and multi-scale context features through structural-aware feature modulation module and inverse aggregation technique. Experimental results show that our method improves accuracy on the Cityscapes (61.2 PQ) and COCO (43.5 PQ) datasets while also demonstrating robustness in challenging simulated real-world complex scenarios faced by robots, such as dirty cameras and rainy conditions. The proposed network promises to help the robot perceive the real scene. In future work, an unsupervised training strategy for the network could be explored to reduce the training cost.

Contextual factors influence food intake at eating occasions in young adults: A mixed effect analysis

Understanding eating behaviours of young adults is crucial for developing strategies to improve their diet quality. Factors that influence eating behaviours are often multifaceted and highly contextual, with little research focused on young adults. This study examined contextual factors influencing food intake at eating occasions among young adults. Participants (n = 675, 18–30 years) recorded dietary intake and contextual characteristics at eating occasions (n = 27,513) using a smartphone food diary (“FoodNow”) over three to four non-consecutive days. Food and beverage intakes were classified into six groups according to the Australian Dietary Guidelines. Eating occasion-level contextual factors, such as consumption location, purchase location, activity, and social context, were recorded via the app, while person-level contextual factors were collected through an online survey. Generalised mixed models with random intercepts were used to investigate effects of contextual factors on food group consumption at meals and snacks. Eating location was associated with vegetable and discretionary food intake; meals that were eaten while in transit were 0.59 serves lower in vegetables (95% CI: −0.93, −0.38; p < 0.001) and 0.31 serves higher in discretionary foods (95% CI: 0.03, 0.54; p = 0.014), compared meals eaten at home. Higher intakes of discretionary foods were observed when friends were present at meals and snacks (0.66 and 0.57 serves, 95% CI: 0.56, 0.88; 0.36, 0.77, p < 0.001). Meals purchased from convenience stores were higher in grain and discretionary foods than meals from other locations (p < 0.001). Majority of person-level factors showed minimal effects on food group consumption. Addressing factors such as purchase location and social contexts during meals and snacks may improve food choices, as small changes at eating occasions can significantly impact overall diet quality over time.

Graph attention automatic encoder based on contrastive learning for domain recognition of spatial transcriptomics

Spatial transcriptomics is an emerging technology that enables the profiling of gene expression in tissues while preserving spatial location information. This innovative approach is anticipated to provide a comprehensive understanding of the spatial distribution of different cells within tissues and facilitate in-depth analysis of tissue structure. To accurately recognize spatial domains from spatial transcriptomics, we have introduced a generalized deep learning method called GAAEST (Graph Attention-based Autoencoder for Spatial Transcriptomics). Our proposed approach effectively integrates both spatial location information and gene expression data from spatial transcriptomics. Specifically, it leverages spatial location details to construct a neighborhood graph and employs a graph attention network-based encoder to embed gene expression information into a spatially informed space. At the same time, to further optimize the learned potential embedding, self-supervised contrastive learning is introduced to capture spatial information at three levels: local, global and contextual feature of spots. Finally, the decoder reconstructs gene expressions, which are then clustered to identify spatial domains with similar expression patterns and spatial proximity. Based on our experiments conducted on multiple datasets, GAAEST consistently outperforms existing state-of-the-art methods. The proposed GAAEST demonstrates excellent capabilities in spatial domain recognition, positioning it as an ideal tool for advancing spatial transcriptomics research.