Chinese Classification Research Articles

Design of Chinese painting style classification model based on multi-layer aggregation CNN

This study delves deeply into exploring the artistic value of traditional Chinese painting (TCP) and aims to bridge the gap between its fundamental characteristics and the realm of human emotions. To achieve this, a novel convolutional neural network (CNN)-based classification model for TCP emotions is proposed. By thoroughly analyzing the distinct emotional mapping relationships inherent in TCP, a comprehensive framework is developed. Notably, emotional feature regions are accurately extracted using image saliency, and a multi-layer aggregation recalibrated emotional feature module is seamlessly integrated into the CNN network structure. This integration strengthens the activation of features that significantly influence TCP emotions. Moreover, a method of multi-category weighted activation localization is skillfully employed to classify Chinese painting emotions within the CMYK (cyan magenta yellow black) color space. The empirical results convincingly demonstrate that our algorithm surpasses existing approaches such as VGG-19, GoogLeNet, ResNet-50, and WSCNet in the task of TCP emotion recognition, achieving an impressive accuracy of 92.36%, the largest error is 0.191. This improvement signifies the advancements made by our model in accurately capturing and understanding the emotional nuances within TCP. By outperforming previous methods, our research contributes to the convergence of multimedia technology and cultural education.

Comparative study of the histology and Micro-CT scan of the goldfish hood

The hood (also known as head growth or wen) is a distinctive feature of goldfish characterized by skin tissue proliferation. The classification of goldfish hood types and the study of the underlying skeletal structure still have substantial gaps in research. This study aimed to investigate the structural differences in goldfish hood tissues using histological and Micro-CT comparisons, based on traditional Chinese classifications. Histological observations revealed two distinct types: the one-piece type (Crown Pearlscale), characterized by a uniformly thickened epidermis, and the fused type (Oranda, Red Cap Oranda, and Ranchu), which showed irregular epidermal extensions into the dermis. Micro-CT scans further demonstrated that hooded goldfish exhibited skull protrusions, bone spurs, and widening of the cranial region. These findings indicate that fibroblast activity, tissue edema, and bone structural changes may be key drivers of hood formation in these goldfish varieties.

Cataloging and the study of ancient Chinese book classification systems

Cataloging and the study of ancient Chinese book classification systems

Hierarchical visual semantic guidance for enhanced relationship recognition in domain knowledge graphs

Multi-modal entity relationship recognition is a crucial foundation for constructing accurate and comprehensive domain knowledge graphs. However, due to the sparse and diverse nature of textual semantic information and the semantic differences exhibited by multi-modal data, existing methods often suffer from issues such as information loss and noise features. In this paper, the Hierarchical Visual Semantic Guidance (HVSG) Network is proposed, which utilizes images to furnish visual semantic information for text, facilitating the identification of implicit relationships between entities. Specifically, salient local instance objects are initially extracted from the global image, followed by the use of a hierarchical visual semantic construction module to establish multi-level visual semantics. The integration of multi-level visual and textual features is then achieved through a visual semantic guidance module. This work constructed a Chinese multi-modal entity relation classification dataset in the domain of unmanned vessels and conducted experiments on two datasets. Compared to state-of-the-art (SOTA) models, HVSG achieved the best performance, with F1 scores improved by 1.05% and 0.58% respectively, indicating that HVSG can offer better performance in exploring implicit relationships between textual entities.

Chinese traditional painting style automatic classification based on dual-channel feature fusion with multi-attention mechanism

Existing classification models for traditional Chinese paintings mostly ignore shallow detail features, which leads to the imprecise classification of styles. To address the above problems, this paper proposes a Chinese traditional painting style automatic classification model based on dual-channel feature fusion with multi-attention mechanism. First, the spatial attention mechanism is introduced to enhance the Swin-Transformer framework to obtain the salient features of Chinese ancient painting images. Second, a dual-channel attention mechanism is constructed to extract global semantic features and local features of Chinese ancient painting images. Finally, the extracted features are fused and categorized based on the softmax classifier. To verify the feasibility and validity of the proposed model, this paper performs simulations on the Chinese painting dataset and compares it with existing algorithms.The average classification accuracy of the proposed model is 90.6[Formula: see text], with an improvement of 3.14[Formula: see text], which is better than the existing model in both visual effects and objective data comparisons.

Impact of comorbid subthreshold depressive symptoms on cancer-related fatigue and complications in adults with leukemia.

Patients not only experience symptoms caused by cancer but also suffer from the accompanying psychological pain. Therefore, these patients do not have high quality of life. According to the World Health Organization, the incidence of leukemia in China in 2020 was 5.1/100000, the mortality rate was 3.3/100000, and the prevalence rate was 16.7/100000. Therefore, it is important to examine the influence of comorbid subthreshold depressive symptoms on leukemia patients. To determine the impact of comorbid subthreshold depressive symptoms on cancer-related fatigue and complications in leukemia patients, thereby providing a basis for early diagnosis and treatment in clinical practice. A questionnaire survey was conducted among leukemia patients admitted to a tertiary hospital in Xi'an, Shaanxi Province, China, from August 2022 to December 2023. Patients with a score > 16 on the Chinese Classification of Mental Disorders (CCMD-3) and a Hamilton Depression Rating Scale score of 8-17 were classified as the subthreshold depressive group (n = 95), while 100 leukemia patients admitted during the same period were classified as the control group. Data were collected using Epidata 3.1 software, and comparisons were made between the two groups regarding general clinical data, the Piper Fatigue Scale (PFS), the Pittsburgh Sleep Quality Index (PSQI), the Numeric Rating Scale for pain assessment, laboratory indicators, and the occurrence of complications. In this survey, 120 leukemia patients with depression were preliminarily screened, 95 patients with subthreshold depression were ultimately selected as the subthreshold depression group, and 100 leukemia patients admitted during the same period were enrolled as the normal group. Comparison of basic clinical data between the two groups revealed no significant differences in age, sex, body mass index, cognitive function, or comorbidity with other chronic diseases. However, there were statistically significant differences in the use of radiotherapy and regular exercise between the two groups (P < 0.05). Comparisons of scales and laboratory indicators revealed no significant differences in albumin or PSQI scores between the two groups, but there were statistically significant differences in pain scores, PSQI scores, PFS scores, hemoglobin levels, and C-reactive protein levels (P < 0.05). Spearman's correlation analysis indicated that cancer-related fatigue was correlated with age, hemoglobin levels, C-reactive protein levels, pain, and regular exercise among leukemia patients with subthreshold depression. Multivariate regression analysis revealed that advanced age, combined radiotherapy, pain, and low hemoglobin levels were risk factors for cancer-related fatigue in leukemia patients with comorbid subthreshold depression, while regular exercise was a protective factor against cancer-related fatigue. Follow-up comparisons revealed a significantly lower overall incidence of complications in the control group (4%) than in the depressive group (24.21%; P < 0.001). Leukemia patients with comorbid subthreshold depressive symptoms experience more severe cancer-related fatigue and a higher incidence of complications. These findings may be related to advanced age, combined radiotherapy, pain, and low hemoglobin levels, while regular exercise may effectively alleviate symptoms.

Modified R-BERT with global semantic information for relation classification task

The objective of the relation classification task is to extract relations between entities. Recent studies have found that R-BERT (Wu and He, 2019) based on pre-trained BERT (Devlin et al., 2019) acquires extremely good results in the relation classification task. However, this method does not take into account the semantic differences between different kinds of entities and global semantic information either. In this paper, we set two different fully connected layers to take into account the semantic difference between subject and object entities. Besides, we build a new module named Concat Module to fully fuse the semantic information among the subject entity vector, object entity vector, and the whole sample sentence representation vector. In addition, we apply the average pooling to acquire a better entity representation of each entity and add the activation operation with a new fully connected layer after our Concat Module. Modifying R-BERT, we propose a new model named BERT with Global Semantic Information (GSR-BERT) for relation classification tasks. We use our approach on two datasets: the SemEval-2010 Task 8 dataset and the Chinese character relationship classification dataset. Our approach achieves a significant improvement over the two datasets. It means that our approach enjoys transferability across different datasets. Furthermore, we prove that these policies we used in our approach also enjoy applicability to named entity recognition task.

Progress and implementing of novel surgical concepts and techniques in patients with resectable gastric cancer liver metastasis

In recent years, advances in surgical techniques and evolving concepts have significantly improved treatment strategies and prognosis for patients with gastric cancer liver metastases. In particular, patients diagnosed with initially resectable gastric cancer liver metastases shows marked improvement in survival. Despite variations in the definition of initially resectable gastric cancer liver metastases among different consensus and guidelines, surgical resection and hepatic physiotherapy are increasingly crucial components of comprehensive treatment. Meanwhile, the advancement of the multidisciplinary team model for diagnosis and treatment, along with the evolution of minimally invasive surgical concepts, offers patients increasingly personalized and less intrusive therapeutic alternatives. According to the Chinese Consensus Classification System for Gastric Cancer Liver Metastasis, the optimal clinical pathway for patients initially diagnosed with resectable gastric cancer liver metastasis involves precise categorization, guided selection of surgical approaches and physiotherapy using the multidisciplinary team model, and consideration of molecular classification. However, the refinement and confirmation of these clinical strategies is still required through high-quality clinical trials.

TCP-RBA: Semi-supervised learning for traditional chinese painting classification with random brushwork augment

Traditional Chinese painting (TCP), culturally significant, reflects China’s rich history and aesthetics. In recent years, TCP classification has shown impressive performance, but obtaining accurate annotations for these tasks is time-consuming and expensive, involving professional art experts. To address this challenge, we present a semi-supervised learning (SSL) method for traditional painting classification, achieving exceptional results even with a limited number of labels. To improve global representation learning, we employ the self-attention-based MobileVit model as the backbone network. Furthermore, We present a data augmentation strategy, Random Brushwork Augment (RBA), which integrates brushwork to enhance the performance. Comparative experiments confirm the effectiveness of TCP-RBA in Chinese painting classification, demonstrating outstanding accuracy of 88.27% on the test dataset, even with only 10 labels, each representing a single class.

CERN生物长期样地背景和植被分类特征本底数据集

This dataset is a comprehensive integration of 95 long-term plots from 22 natural ecosystem stations. We first sorted out the background information such as the geographical location, establishment time, area and representativeness of the long-term plot of ecological station. Drawing from the survey data of plant communities at the beginning of the plot establishment and adhering to the latest Chinese vegetation classification system, we then meticulously delineated the vegetation formation group, vegetation formation and vegetation subformation of each sample plot’s plant community. Based on the list of dominant species found in each sample plot, we identified and named the alliance and association of the vegetation in the plot. The construction of this dataset went through multiple rounds of expert review-station verification-expert review-station acceptance, etc. The dataset comprises station code, station name, plot code, plot name, plot type, plot representativeness, geographical location, altitude, plot area and shape, plot establishment time and design service life, vegetation formation group, vegetation formation, vegetation subformation, alliance, association, dominant species and remarks. It is expected to provide data support for the studies on flora, vegetation resource distribution, and biodiversity conservation.

Association between BMI and osteoporotic fractures at different sites in Chinese women: a case-control retrospective study in Changsha

BackgroundOsteoporotic fractures are a growing problem in an aging society. The association between body mass index (BMI) and osteoporotic fractures varies by fracture site and ethnicity. Limited knowledge exists regarding this association in native Chinese, particularly utilizing local databases as reference sources.ObjectiveTo investigate the association between BMI and osteoporotic fractures at different sites in Chinese women.MethodsThree thousand ninety-eight female patients with radiographic fractures and 3098 age- and sex-matched healthy controls without fractures were included in the study. Both of them underwent assessment using dual-energy X-ray absorptiometry (DXA), with BMD measurements calculated using our own BMD reference database. Participants were classified into underweight (BMI < 18.5 kg/m2), normal weight (18.5 ≤ BMI < 24.0 kg/m2), overweight (24 ≤ BMI < 28 kg/m2) and obese (BMI ≥ 28 kg/m2) according to the Chinese BMI classification standard.ResultsThere were 2296 (74.1%) vertebral fractures, 374 (12.1%) femoral neck fractures, and 428 (13.8%) other types of fractures in the case group. Bone mineral density (BMD) was almost lower in the fracture groups compared to the control groups (p = 0.048 to < 0.001). Compared with normal weight, underweight had a protective effect on total [odds ratio (OR) = 0.61; 95% confidence interval (CI), 0.49 –0.75; P< 0.001], and lumbar fractures (OR = 0.52; 95% CI, 0.41 – 0.67; P < 0.001), while obesity was associated with an increased risk for total (OR = 2.26; 95% CI, 1.85 – 2.76; P < 0.001), lumbar (OR = 2.17; 95% CI, 1.72 – 2.73; P < 0.001), and femoral neck fractures (OR = 4.08; 95% CI, 2.18 – 7.63; P < 0.001). Non-linear associations were observed between BMI and fractures: A J-curve for total, lumbar, and femoral neck fractures, and no statistical change for other types of fractures. Underweight was found to be a risk factor for other types of fracturess after adjusting for BMD (OR = 2.29; 95% CI, 1.09 – 4.80; P < 0.001). Osteoporosis and osteopenia were identified as risk factors for almost all sites of fracture when compared to normal bone mass.ConclusionsUnderweight has a protective effect on total and lumbar spine fractures in Chinese women, while obesity poses a risk factor for total, lumbar, and femoral neck fractures. The effect of BMI on fractures may be mainly mediated by BMD.

MGMFN: Multi-graph and MLP-mixer fusion network for Chinese social network sentiment classification

MGMFN: Multi-graph and MLP-mixer fusion network for Chinese social network sentiment classification

Creation of Deep Learning Scenarios in the Network Teaching of Physical Education Technical Courses

The network teaching evaluation of sports professional technical courses has positive significance for the sustainable development of education. And how to establish an effective evaluation model is the key part. The research introduces the creation of in-depth learning scenarios (LS) into the network teaching of sports professional technical courses, and then constructs a new network teaching mode of sports professional technical courses. The Particle Swarm Optimization algorithm - Attention- Long Short-Term Memory network (PSO-Attention-LSTM) Chinese Emotion Classification Model (ECM) is constructed to classify the online evaluation text to realize the evaluation of online teaching. This model combines the improved PSO, Attention and LSTM classification models. The optimal number of hidden layer nodes for LSTM model is about 100, and the optimal data size for batch processing is 25. The overall error rate of online teaching teachers of male and female sports professional skill courses is 10.1%, and the overall error rate of online teaching teachers of general and advanced sports professional skill courses is 12.8%. The application effect of the creation of in-depth learning scene in the network teaching of physical education technical courses is shown. When the classification threshold is 0.6 and 0.8 respectively, the AUC of PSO-Attention-LSTM Chinese ECM is 0.821 and 0.809 respectively. The research institute has put forward that the online teaching platform of sports professional technical courses has extremely high practical application value and has been unanimously praised by network users.

Hierarchical multi-instance multi-label learning for Chinese patent text classification

To further enhance the accuracy of the Chinese patent classification, this paper proposes a model, based on the patent structure and takes the patent claim as subjects, with multi-instance multi-label learning as the main method. Firstly, the patent claims are divided into multiple independent texts using the sequence number as the splitting token. For each patent, multiple claims are regarded as multiple instances, and the corresponding IPCs serve as its multiple labels. Next, the concept of secondary_label is introduced following the composition rules of IPC, and the relationships between instances and multiple secondary_labels are mined through the construction of fully-connected layers. To capture more comprehensive semantic information of instances, BIGRU and self-attention are employed to enhance semantics and reduce information loss during the training process. Finally, the max-pooling operations are utilised to obtain the predicted categories of patents based on capturing the relationships between instances and different hierarchical labels. Experimental results on the '2017 Chinese patent dataset' demonstrate that the multi-instance multi-label approach can effectively mine deeper relationships between patents and labels in classification tasks. As a result, our model significantly improves the accuracy of patent text classification.

On the Origin of the Symbol “Dragon” from the Zoonym “Crocodile” in the Cultures of Ancient Southern China and Vietnam

The author explores the origin of the symbol “dragon” in the cultures of ancient Southern China and Vietnam from the “crocodile”, puts forward the hypothesis that the ancient ethnonym and toponym “Yue” 越, (modern Vietnam) come from the Chinese zoonym ‘crocodile’ 鳄鱼 èyú. Other sources are the words “yue” 岳 ‘mountain’, “yue” 钺 ‘great battle axe’ and “yue” 月 ‘moon’. The Vietnamese word âu ‘seagull’, isomorphic to the Chinese word 鸥 ‘ōu’, is also involved in the semiotic analysis. Correlation of the Chinese letter from “jiaguwen” to “kaishu” (with the mention of the Vietnamese “ty-nom” and the letter of the state “Dian”) with historical and modern reality contributes to the objectification of the Chinese-Vietnamese world, the possibility of studying it using philological knowledge. The article also provides a comparison of the Chinese classification of nature by Zhou Qu-fei with the system of Carl Linnaeus and the earlier classification of Conrad Gesner, and develops the idea of the economy of nature.

Long short-term memory (LSTM)-based news classification model.

In this study, we used unidirectional and bidirectional long short-term memory (LSTM) deep learning networks for Chinese news classification and characterized the effects of contextual information on text classification, achieving a high level of accuracy. A Chinese glossary was created using jieba-a word segmentation tool-stop-word removal, and word frequency analysis. Next, word2vec was used to map the processed words into word vectors, creating a convenient lookup table for word vectors that could be used as feature inputs for the LSTM model. A bidirectional LSTM (BiLSTM) network was used for feature extraction from word vectors to facilitate the transfer of information in both the backward and forward directions to the hidden layer. Subsequently, an LSTM network was used to perform feature integration on all the outputs of the BiLSTM network, with the output from the last layer of the LSTM being treated as the mapping of the text into a feature vector. The output feature vectors were then connected to a fully connected layer to construct a feature classifier using the integrated features, finally classifying the news articles. The hyperparameters of the model were optimized based on the loss between the true and predicted values using the adaptive moment estimation (Adam) optimizer. Additionally, multiple dropout layers were added to the model to reduce overfitting. As text classification models for Chinese news articles, the Bi-LSTM and unidirectional LSTM models obtained f1-scores of 94.15% and 93.16%, respectively, with the former outperforming the latter in terms of feature extraction.

Multimodal Chinese Agricultural News Classification Method Based on Interactive Attention

Multimodal Chinese Agricultural News Classification Method Based on Interactive Attention

Revised Proposed Classifications for Typical Anthropogenic Soils in China

As global industrialization and its associated anthropogenic activities rapidly increase, so too does the areal extent of human-altered soils. The soil classification framework must incorporate the classification schemes of these disturbed soils in order to remediate land. Soil Taxonomy (ST) and the World Reference Base for Soil Resources (WRB) are the most widely used soil classification systems in the world. In this study, 15 typical anthropogenic soil profiles with engineering and technical characteristics from China were selected for a classification study. The aim of this study was to clarify the classifications of these soil profiles in the ST, WRB, Chinese Soil Taxonomy (CST), and Geogenetic Soil Classification of China (GSCC), and make references accordingly. The results showed that the WRB can classify these soils as Technosols in the first level. ST can classify most of these soils as subgroups within the different great groups under Entisols, as well as the Human-Altered and Human-Transported Material classes within the soil family differentiae. For Chinese soil classification schemes, there is a large loophole in the CST regarding the classification of anthropogenic soils. Many anthropogenic soils cannot reflect these soils’ artificial and technical naming characteristics. For the CST, revised proposals based on the WRB and ST were proposed. Based on the artificial disturbance of soil layers, a manufactured layer was added to the diagnostic basis. The Artificalic Anthrosols suborder was added, and it can be divided into the Transporti-Artificalic Anthrosols and Alteri-Artificalic Anthrosols groups. The subgroups were defined by their levels of contamination, imperviousness, and artifact content. The “T” layer symbol was added to soil profile descriptions to reflect a specific layer about technology, such as A, E, B, and C, to reflect the main genetic horizons. This revised classification system is proposed for inclusion in the revised CST to account for the very large and expanding extent of disturbed soils in China and to remain current with other global soil taxonomy systems.

A new empirical model for the undrained shear strength of undisturbed marine fine-grained soil based on piezocone penetration tests

Simple, precise, and economical evaluation of the undrained shear strength Su of undisturbed marine soil can facilitate the construction of offshore projects. In this study, piezocone penetration tests (CPTu) were conducted at the site of a wind power project, and consolidated undrained triaxial tests were performed on specimens of undisturbed marine fine-grained soil taken from the site. Analysis shows that the Eslami–Fellenius and Chinese soil classification methods based on CPTu data are more applicable for marine soils. The test results indicate that the stress–strain curve exhibits strain softening if the clay content is less than 30 %, otherwise it exhibits either strain hardening or strain stability. Furthermore, the specimens exhibit two deformation failure modes: softening contraction and hardening dilatancy. From the correlation between the consolidated undrained triaxial test data and the CPTu data, Su can be determined simply via an explicit expression involving the measured pore pressure behind the cone (u2), the corrected cone tip resistance (qt), and the soil depth H, and this expression is significantly more precise than the existing empirical formulas. The proposed method offers a sound basis for designing the foundations of marine engineering and infrastructure projects in different sea region.

A quantitative assessment method for the correlation between rock mass basic quality and Protodyakonov coefficient

With the implementation of the “Belt and Road” initiative, tunnel projects were designed in accordance with Russian design specifications and constructed by Chinese companies in countries along the route. Design companies and construction companies use different rock classification methods, which will increase the safety risk and cost of the tunnels. Therefore, it is necessary to study the correlation between Chinese engineering rock mass classification and Russian rock mass classification. The goal was to establish the correlation between rock mass basic quality (BQ) and Protodyakonov coefficient (f) for rock mass classification. Firstly, based on the equivalence principle of uniaxial compressive strength (UCS), the relationship between BQ and f containing the velocity index of rock mass (Kv) was established by comparing the relationship between f and BQ in the empirical formula of UCS. Secondly, through the equivalent principle of volumetric joint count of rock mass (Jv), we compared the empirical formula between the intactness index of Kv and rock quality designation (RQD) and conducted linear regression analysis on a large amount of data. The relationship between Kv and RQD was established. Finally, the relationship between BQ and f is quantitatively expressed and a classification table corresponding to the Russian rock classification and the Chinese engineering rock classification is obtained. The method is applied to Kyrgyzstan mountain tunnels.