Powerful Representation Research Articles

ANALYSIS OF THE CONCLUSION OF THE PLENUM OF THE CONSTITUTIONAL COURT IN RELATION TO THE VIOLATION OF THE CONSTITUTION BY THE PRESIDENT OF GEORGIA AND INTERNATIONAL PRACTICE

From 2013 until the present, the dispute on the separation of competencies between the leaders of the state government has been ongoing in political life, as well as on an international scale in the field of foreign relations. The field of foreign relations is of national importance and is directly related to the state’s image and foreign policy. The conclusion of the international treaties, as well as the appointment of the ambassadors and representation of the state, has become the subject of dispute at various times. State representation by the President in foreign relations has led to a Constitutional Reference submitted by the Parliament of Georgia to the Constitutional Court of Georgia, alleging a violation of the Constitution by the President for the first time in the country’s history. As a result, according to the conclusion of the Constitutional Court of Georgia dated October 16, 2023, President Ms. Zurabishvili violated the Constitution of Georgia by exercising representation powers without the consent of the Government. Based on the above, this thesis will analyze the meaning of the representation power in foreign relations, the President’s constitutional status, and the President’s role in foreign relations, acting as the socalled “neutral arbitrator”. This thesis will include a discussion on the roles of the Parliament and the Constitutional Court of Georgia in the political-legal process of Presidential impeachment proceedings. It will also examine the conclusions and decisions of Constitutional Courts in various states regarding alleged constitutional violations by Presidents and disputes over competence.

Formation of Social Ties in P. I. Potemkin's Embassy to France (1680–1682)

Based on the extant complex of Russian and French sources, this article analyses the mechanisms of the formation of social ties of the Russian embassy in France in 1681. The embassy had both foreign policy and diplomatic significance for the representation of tsarist power and Russian culture in France: the envoys had not only to inform Louis XIV about the events in Eastern Europe, but also to get an overview of the French position on the preparation of the Perpetual Peace of Russia with the Polish-Lithuanian Commonwealth, the conclusion of the Russian-French trade treaty, and to ensure the recruitment of foreign specialists. The author examines the formation of relations in the context of the new diplomatic history, paying special attention to the problems of power representation, the personalities of diplomats and the cultural side of their mission. For the first time in historiography, the article characterises the whole complex of archival documents about the embassy, the peculiarities of the formation of social ties of the embassy participants necessary for the successful implementation of the tasks of the diplomatic mission. The study is based on French court reports on ceremonies, notes of courtiers, internal correspondence of Russian and French diplomatic departments, materials of the Parisian periodical press, and reports of Russian envoys. Conclusions are drawn on the peculiarities of the strategies of the envoy P. I. Potemkin in establishing social relations, their significance for the development of Russian diplomacy, foreign policy, and cultural contacts with Western Europe, as well as on the peculiarities of the reflection of these relations in the surviving sources.

EEG Ictal Power Dynamics, Function-Structure Associations, and Epilepsy Surgical Outcomes.

Postoperative seizure control in drug-resistant temporal lobe epilepsy (TLE) remains variable, and the causes for this variability are not well understood. One contributing factor could be the extensive spread of synchronized ictal activity across networks. Our study used novel quantifiable assessments from intracranial EEG (iEEG) to test this hypothesis and investigated how the spread of seizures is determined by underlying structural network topological properties. We evaluated iEEG data from 157 seizures in 27 patients with TLE: 100 seizures from 17 patients with postoperative seizure control (Engel score I) vs 57 seizures from 10 patients with unfavorable surgical outcomes (Engel score II-IV). We introduced a quantifiable method to measure seizure power dynamics within anatomical regions, refining existing seizure imaging frameworks and minimizing reliance on subjective human decision-making. Time-frequency power representations were obtained in 6 frequency bands ranging from theta to gamma. Ictal power spectrums were normalized against a baseline clip taken at least 6 hours away from ictal events. Electrodes' time-frequency power spectrums were then mapped onto individual T1-weighted MRIs and grouped based on a standard brain atlas. We compared spatiotemporal dynamics for seizures between groups with favorable and unfavorable surgical outcomes. This comparison included examining the range of activated brain regions and the spreading rate of ictal activities. We then evaluated whether regional iEEG power values were a function of fractional anisotropy (FA) from diffusion tensor imaging across regions over time. Seizures from patients with unfavorable outcomes exhibited significantly higher maximum activation sizes in various frequency bands. Notably, we provided quantifiable evidence that in seizures associated with unfavorable surgical outcomes, the spread of beta-band power across brain regions is significantly faster, detectable as early as the first second after seizure onset. There was a significant correlation between beta power during seizures and FA in the corresponding areas, particularly in the unfavorable outcome group. Our findings further suggest that integrating structural and functional features could improve the prediction of epilepsy surgical outcomes. Our findings suggest that ictal iEEG power dynamics and the structural-functional relationship are mechanistic factors associated with surgical outcomes in TLE.

Motion-guided spatiotemporal multitask feature discrimination for self-supervised video representation learning

Powerful self-supervised representation models are able to step out of the traditional supervised paradigm and rely merely on unlabeled data to achieve a deep understanding of visual semantic features. However, previous approaches may suffer from background scene bias, making it difficult to perform a comprehensive exploration of video spatiotemporal structure. To address this challenge, this paper proposes a self-supervised video representation learning framework of motion-guided spatiotemporal multitask feature discrimination (MSMFD). The method mainly utilizes the consistency of motion cues between different views to guide the model for spatial and temporal feature similarity discrimination. Specifically, the model first selects video clips with large motion amplitudes based on the collected optical flow maps. Subsequently, the model introduces an instance discrimination task for overall spatiotemporal structure perception of the video, while a shuffled triplet and an augmented quadruple task are created to further enhance the exploration of intraframe sequence order and local spatial fine-grained. Furthermore, we propose joint motion alignment of spatial, temporal, and spatiotemporal dimensions under different views as a powerful compensation for acquiring motion features. Experimental results demonstrate that our self-supervised method is effective for learning video representations and achieves competitive performance in action recognition and video retrieval tasks compared to other state-of-the-art methods.

A class-aware representation refinement framework for graph classification

Graph Neural Networks (GNNs) are widely used for graph representation learning. Despite its prevalence, GNN suffers from two drawbacks in the graph classification task, the neglect of graph-level relationships, and the generalization issue. Each graph is treated separately in GNN message passing/graph pooling, and existing methods to address overfitting operate on each individual graph. This makes the graph representations learnt less effective in the downstream classification. In this paper, we propose a Class-Aware Representation rEfinement (CARE) framework for the task of graph classification. CARE computes simple yet powerful class representations and injects them to steer the learning of graph representations towards better class separability. CARE is a plug-and-play framework that is highly flexible and able to incorporate arbitrary GNN backbones without significantly increasing the computational cost. We also theoretically prove that CARE has a better generalization upper bound than its GNN backbone through Vapnik-Chervonenkis (VC) dimension analysis. Our extensive experiments with 11 well-known GNN backbones on 9 benchmark datasets validate the superiority and effectiveness of CARE over its GNN counterparts.

Markov conditions and factorization in logical credal networks

We examine the recently proposed language of Logical Credal Networks, a powerful representation formalism that combines probabilities and logic. In particular we investigate the consequences of distinct Markov conditions upon their underlying semantics. We introduce the notion of structure for a Logical Credal Network and show that a structure without directed cycles leads to a well-known factorization result. For networks with directed cycles, we discuss the differences between Markov conditions, factorization results, and specification requirements. We consider several scenarios in causal reasoning that can be tackled by the formalism, in particular looking at partial identifiability and cycles.

Мамлюкская сакральная топография: стратегия и практика репрезентации власти

Султанат мамлюков был образован рабами-воинами в Египте в середине XIII в. и просуществовал более 250 лет. Разгром армии Людовика IX в 1250 г. и победа при ʿАйн Джалуте в 1260 г. дали мамлюкам право считать себя подлинными владыками одной из самых могущественных держав Ближнего Востока. Период правления мамлюкских султанов (1250–1517) был отмечен строительством многочисленных и весьма заметных до сих пор в современном ландшафте Каира, который был столицей Султаната, погребальных сооружений – мавзолеев. Вокруг них формировались большие или малые религиозные комплексы, которые включали мечеть или медресе, суфийскую обитель, библиотеку. Мавзолеи становились местом сакрального притяжения для самих мамлюков, но вместе с тем участвовали в формировании пространства, имевшего сакральный характер для всего населения Египта. Особое значение приобрели погребальные религиозные комплексы в период правления черкесских султанов (1382–1517). Свергнув своих предшественников Бахритов-Калаунидов (1250–1382) и отступив от династийного принципа престолонаследия, они нуждались в дополнительных источниках укрепления легитимности власти и активно поддерживали деятельность исламских религиозных и образовательных учреждений, а величественные мавзолеи как ориентиры памяти и незыблемые городские доминанты отчасти компенсировали недостаток стабильности в условиях нединастийного перехода власти. The Mamluk Sultanate was formed by slave warriors in Egypt in the mid-13th century and existed for more than 250 years. The defeat of the army of Louis IX in 1250 and the victory at Ain Jalut in 1260 gave the Mamluks the right to consider themselves the true rulers of one of the most powerful powers in the Middle East. The period of the reign of the Mamluk sultans (1250–1517) was marked by the construction of numerous mausoleums that are still very noticeable in the modern landscape of Cairo, which was the capital of the Sultanate. Large or small religious complexes were formed around them, which included a mosque or madrasah, a Sufi monastery, or a library. Mausoleums became a place of sacred attraction for the Mamluks themselves, but, at the same time, they participated in the formation of a space that had a sacred character for the entire population of Egypt. Funeral religious complexes acquired particular importance during the reign of the Circassian Mamluk sultans (1382–1517). Having overthrown their predecessors, the Bahri-Qalawnids (1250–1382) and retreated from the dynastic principle of succession to the throne, the Circassian Mamluks needed additional sources of strengthening the legitimacy of power and actively supported the activities of Islamic religious and educational institutions, and the majestic mausoleums as landmarks of memory and unshakable urban dominants partly compensated for the lack of stability in conditions of non-dynasty transition of power.

Detecting abnormal logins by discovering anomalous links via graph transformers

Anomalous authentications are a critical indicator of advanced persistent threats (APTs), in which adversaries exploit network vulnerabilities to gain unauthorized access and move stealthily between devices using stolen credentials. As the set of interactions between entities in a network essentially forms graph-structured data, state-of-the-art algorithms such as graph neural networks (GNNs) can be used to detect anomalous interactions that may indicate an ongoing attack. However, the success of detecting anomalous authentications using GNNs is conditioned on the representational power and performance of those models. A crucial problem is how to aggregate the node embeddings so that the GNN can better represent the network topology. Existing graph neural networks traditionally use simple functions (e.g., sum, max, mean) on the node embeddings to preserve permutation invariance and achieve consistent node representations. However, we argue that an effective aggregation of node features into a graph-level representation cannot be achieved through simple sum or mean operations. In this work, we propose a residual soft-attention scheme that facilitates the aggregation of node representations through a weighted sum, resulting in enhanced node representations and improved filtration of irrelevant information. Experimental results on three relevant datasets have shown the proposed method can detect abnormal authentications with lower false positives than competitors.

Pipelining Semantic Expansion and Noise Filtering for Sentiment Analysis of Short Documents – CluSent Method

The challenge of constructing effective sentiment models is exacerbated by a lack of sufficient information, particularly in short texts. Enhancing short texts with semantic relationships becomes crucial for capturing affective nuances and improving model efficacy, albeit with the potential drawback of introducing noise. This article introduces a novel approach, CluSent, designed for customized dataset-oriented sentiment analysis. CluSent capitalizes on the CluWords concept, a proposed powerful representation of semantically related words. To address the issues of information scarcity and noise, CluSent addresses these challenges: (i) leveraging the semantic neighborhood of pre-trained word embedding representations to enrich document representation and (ii) introducing dataset-specific filtering and weighting mechanisms to manage noise. These mechanisms utilize part-of-speech and polarity/intensity information from lexicons. In an extensive experimental evaluation spanning 19 datasets and five state-of-the-art baselines, including modern transformer architectures, CluSent emerged as the superior method in the majority of scenarios (28 out of 38 possibilities), demonstrating noteworthy performance gains of up to 14% over the strongest baselines.

Fragmentation or Cooperation: Consociational Structures in Israel & Germany

In the past several years, Israeli politics have entered an unprecedented era of divisiveness and vitriol. After five elections in four years and crisis after crisis, an analysis of this decline of institutions is crucial to understanding how democratic trajectories function, especially in societies as diverse as Israel’s. Germany’s contrasting maintenance of many of Israel’s lost democratic habits serves as an intriguing prism through which to view Israel. What do the two countries’ systems do differently? This paper will argue that Israel’s recent turn and Germany’s stability are the result of the fundamental flaws of the Israeli political system’s structure as it relates to Arend Lijphart’s consociational model for governing divided societies. Even if not in a purely consociational system, are there features of Germany’s political structure Israel could implement? Three of Lijphart’s tenets will be dissected with respect to these differences: segmental autonomy, proportional representation, and balance of power.

Enhanced analysis of tabular data through Multi-representation DeepInsight

Tabular data analysis is a critical task in various domains, enabling us to uncover valuable insights from structured datasets. While traditional machine learning methods can be used for feature engineering and dimensionality reduction, they often struggle to capture the intricate relationships and dependencies within real-world datasets. In this paper, we present Multi-representation DeepInsight (MRep-DeepInsight), a novel extension of the DeepInsight method designed to enhance the analysis of tabular data. By generating multiple representations of samples using diverse feature extraction techniques, our approach is able to capture a broader range of features and reveal deeper insights. We demonstrate the effectiveness of MRep-DeepInsight on single-cell datasets, Alzheimer's data, and artificial data, showcasing an improved accuracy over the original DeepInsight approach and machine learning methods like random forest, XGBoost, LightGBM, FT-Transformer and L2-regularized logistic regression. Our results highlight the value of incorporating multiple representations for robust and accurate tabular data analysis. By leveraging the power of diverse representations, MRep-DeepInsight offers a promising new avenue for advancing decision-making and scientific discovery across a wide range of fields.

DualTrans: A Novel Glioma Segmentation Framework Based on a Dual-Path Encoder Network and Multi-View Dynamic Fusion Model

Segmentation methods based on convolutional neural networks (CNN) have achieved remarkable results in the field of medical image segmentation due to their powerful representation capabilities. However, for brain-tumor segmentation, owing to the significant variations in shape, texture, and location, traditional convolutional neural networks (CNNs) with limited convolutional kernel-receptive fields struggle to model explicit long-range (global) dependencies, thereby restricting segmentation accuracy and making it difficult to accurately identify tumor boundaries in medical imaging. As a result, researchers have introduced the Swin Transformer, which has the capability to model long-distance dependencies, into the field of brain-tumor segmentation, offering unique advantages in the global modeling and semantic interaction of remote information. However, due to the high computational complexity of the Swin Transformer and its reliance on large-scale pretraining, it faces constraints when processing large-scale medical images. Therefore, this study addresses this issue by proposing a smaller network, consisting of a dual-encoder network, which also resolves the instability issue that arises in the training process of large-scale visual models with the Swin Transformer, where activation values of residual units accumulate layer by layer, leading to a significant increase in differences in activation amplitudes across layers and causing model instability. The results of the experimental validation using real data show that our dual-encoder network has achieved significant performance improvements, and it also demonstrates a strong appeal in reducing computational complexity.

Enhancing Livestock Detection: An Efficient Model Based on YOLOv8

Maintaining a harmonious balance between grassland ecology and local economic development necessitates effective management of livestock resources. Traditional approaches have proven inefficient, highlighting an urgent need for intelligent solutions. Accurate identification of livestock targets is pivotal for precise livestock farming management. However, the You Only Look Once version 8 (YOLOv8) model exhibits limitations in accuracy when confronted with complex backgrounds and densely clustered targets. To address these challenges, this study proposes an optimized CCS-YOLOv8 (Comprehensive Contextual Sensing YOLOv8) model. First, we curated a comprehensive livestock detection dataset encompassing the Qinghai region. Second, the YOLOv8n model underwent three key enhancements: (1) incorporating a Convolutional Block Attention Module (CBAM) to accentuate salient image information, thereby boosting feature representational power; (2) integrating a Content-Aware ReAssembly of FEatures (CARAFE) operator to mitigate irrelevant interference, improving the integrity and accuracy of feature extraction; and (3) introducing a dedicated small object detection layer to capture finer livestock details, enhancing the recognition of smaller targets. Experimental results on our dataset demonstrate the CCS-YOLOv8 model’s superior performance, achieving 84.1% precision, 82.2% recall, 84.4% mAP@0.5, 60.3% mAP@0.75, 53.6% mAP@0.5:0.95, and 83.1% F1-score. These metrics reflect substantial improvements of 1.1%, 7.9%, 5.8%, 6.6%, 4.8%, and 4.7%, respectively, over the baseline model. Compared to mainstream object detection models, CCS-YOLOv8 strikes an optimal balance between accuracy and real-time processing capability. Its robustness is further validated on the VisDrone2019 dataset. The CCS-YOLOv8 model enables rapid and accurate identification of livestock age groups and species, effectively overcoming the challenges posed by complex grassland backgrounds and densely clustered targets. It offers a novel strategy for precise livestock population management and overgrazing prevention, aligning seamlessly with the demands of modern precision livestock farming. Moreover, it promotes local environmental conservation and fosters sustainable development within the livestock industry.

Learning Robust Representations of Tonic-Clonic Seizures With Cyclic Transformer.

Tonic-clonic seizures (TCSs) pose a significant risk for sudden unexpected death in epilepsy (SUDEP). Previous research has highlighted the potential of multimodal wearable seizure detection systems in accurately detecting TCSs through continuous monitoring, enabling timely alarms and potentially preventing SUDEP. However, such multimodal systems carry a higher risk of sensor malfunction. In this paper, we propose a cyclic transformer approach to address these challenges. The cyclic transformer learns a robust representation by performing circular modal translations between the source and target modalities. It leverages back-translation as regularization technique to enhance the discriminative power of the learned representation. Notably, the proposed cyclic transformer is trained on paired multimodal data but requires only a single source modality during deployment. This characteristic ensures the robustness of the cyclic transformer to perturbations or missing information in the target modality. Experimental results demonstrate that the proposed cyclic transformer achieves competitive performance compared with existing multimodal systems. While both approaches were trained using EEG and EMG data, the cyclic transformer exclusively employs EEG data for testing, diverging from the state-of-the-art's utilization of both EEG and EMG data during test. This showcases the effectiveness of the cyclic transformer in multimodal TCSs detection, offering a promising approach for enhancing the accuracy and robustness of seizure detection systems while mitigating the risks associated with sensor malfunction.

Self-supervised learning with automatic data augmentation for enhancing representation

Self-supervised learning has become an increasingly popular method for learning effective representations from unlabeled data. One prominent approach in self-supervised learning is contrastive learning, which trains models to distinguish between similar and dissimilar sample pairs by pulling similar pairs closer and pushing dissimilar pairs farther apart. The key to the success of contrastive learning lies in the quality of the data augmentation, which increases the diversity of the data and helps the model learn more powerful and generalizable representations. While many studies have emphasized the importance of data augmentation, however, most of them rely on human-crafted augmentation strategies. In this paper, we propose a novel method, Self Augmentation on Contrastive Learning with Clustering (SACL), searching for the optimal data augmentation policy automatically using Bayesian optimization and clustering. The proposed approach overcomes the limitations of relying on domain knowledge and avoids the high costs associated with manually designing data augmentation rules. It automatically captures informative and useful features within the data by exploring augmentation policies. We demonstrate that the proposed method surpasses existing approaches that rely on manually designed augmentation rules. Our experiments show SACL outperforms manual strategies, achieving a performance improvement of 1.68% and 1.57% over MoCo v2 on the CIFAR10 and SVHN datasets, respectively.

Intelligent quality control of traditional chinese medical tongue diagnosis images based on deep learning.

Computer-aided tongue and face diagnosis technology can make Traditional Chinese Medicine (TCM) more standardized, objective and quantified. However, many tongue images collected by the instrument may not meet the standard in clinical applications, which affects the subsequent quantitative analysis. The common tongue diagnosis instrument cannot determine whether the patient has fully extended the tongue or collected the face. This paper proposes an image quality control algorithm based on deep learning to verify the eligibility of TCM tongue diagnosis images. We firstly gathered enough images and categorized them into five states. Secondly, we preprocessed the training images. Thirdly, we built a ResNet34 model and trained it by the transfer learning method. Finally, we input the test images into the trained model and automatically filter out unqualified images and point out the reasons. Experimental results show that the model's quality control accuracy rate of the test dataset is as high as 97.06%. Our methods have the strong discriminative power of the learned representation. Compared with previous studies, it can guarantee subsequent tongue image processing. Our methods can guarantee the subsequent quantitative analysis of tongue shape, tongue state, tongue spirit, and facial complexion.

Dialogues between Past and Present? Modern Art, Contemporary Art Practice, and Ancient Egypt in the Museum

Whenever twentieth-century modern art or new contemporary artworks are included amongst displays of ancient Egypt, press statements often assert that such juxtapositions are ‘surprising’, ‘innovative’, and ‘fresh’, celebrating the external perspective they bring to such collections. But contemporary art’s relationship with museums and other disciplines needs to be understood in a longer-term perspective. Pairings of twentieth- and twenty-first-century artistic works with objects of antiquity is an activity that has been undertaken for more than a century in what has been a relatively long period of mutually reinforcing influences between modern/contemporary art, museum display, the art market, and Egyptian heritage. Together, they have decontextualised ancient Egyptian culture and shaped the language and perspectives of scholars, curators, and artists. In this paper, rather than considering how artists have been inspired by ancient Egypt, I will give a few examples of how more recent art practices from the late nineteenth century onwards have impacted the language and discourse of Egyptology and its museum representation. Then, using more recent artist engagements with the British Museum, I argue for greater interdisciplinary dialogues between artists and Egyptologists, as both take more critical stances towards research that recontextualises the power and agency of collections, representation, and knowledge production.

Construct representation and predictive validity of integrated writing tasks: A study on the writing component of the Duolingo English Test

This study examined whether two integrated reading-to-write tasks could broaden the construct representation of the writing component of Duolingo English Test (DET). It also verified whether they could enhance DET’s predictive power of English academic writing in universities. The tasks were (1) writing a summary based on two source texts and (2) writing a reading-to-write essay based on five texts. Both were given to a sample (N = 204) of undergraduates from Hong Kong. Each participant also submitted an academic assignment written for the assessment of a disciplinary course. Three professional raters double-marked all writing samples against detailed analytical rubrics. Raw scores were first processed using Multi-Faceted Rasch Measurement to estimate inter- and intra-rater consistency and generate adjusted (fair) measures. Based on these measures, descriptive analyses, sequential multiple regression, and Structural Equation Modeling were conducted (in that order). The analyses verified the writing tasks’ underlying component constructs and assessed their relative contributions to the overall integrated writing scores. Both tasks were found to contribute to DET’s construct representation and add moderate predictive power to the domain performance. The findings, along with their practical implications, are discussed, especially regarding the complex relations between construct representation and predictive validity.

Impact of Fashion on the Body Image of Adult of Age Group 18-60 Years Living in Mumbai

This paper explores the complex connection between women’s body image and fashion, with an emphasis on modern influences and ramifications. It investigates how the fashion industry affects women’s perceptions of their bodies and defines societal standards of beauty through an extensive analysis of scholarly literature and media sources. The paper discusses the self perception of women’s body image and their several methods to alter their appearance. The impact that fashion industry, social media, and celebrity culture have in encouraging women to have unsatisfactory bodies and unattainable body ideals is examined. The article also addresses the rise of body-positive movements and diverse fashion representation, emphasizing how these developments have the ability to subvert conventional notions of beauty and advance acceptance of a range of body shapes. This paper adds to a deeper understanding of the complex interactions between fashion, media, and self-perception by examining both the positive and negative effects of fashion on women’s body image. In the end, it advocates for more inclusive and powerful representations of women in the fashion industry.

Anticipating Trends and Innovations for A Sustainable Automotive Industrial Revolution

Graph classification aims to predict the label associated with a graph and is an important graph analytic task with widespread applications. Recently, graph neural networks (GNNs) have achieved state-of-the-art results on purely supervised graph classification by virtue of the powerful representation ability of neural networks. However, almost all of them ignore the fact that graph classification usually lacks reasonably sufficient labelled data in practical scenarios due to the inherent labelling difficulty caused by the high complexity of graph data. The existing semi-supervised GNNs typically focus on the task of node classification and are incapable of dealing with graph classification. To tackle the challenging but practically useful scenario, we propose a novel and general semi-supervised GNN framework for graph classification, which takes full advantage of a slight amount of labelled graphs and abundant unlabelled graph data. In our framework, we train two GNNs as complementary views for collaboratively learning high-quality classifiers using both labelled and unlabelled graphs. To further exploit the view itself, we constantly select pseudo-labelled graph examples with high confidence from its own view for enlarging the labelled graph dataset and enhancing predictions on graphs. Furthermore, the proposed framework is investigated on two specific implementation regimes with a few labelled graphs and the extremely few labelled graphs, respectively. Extensive experimental results demonstrate the effectiveness of our proposed semi-supervised GNN framework for graph classification on several benchmark datasets