Search

Natural language generation (NLG) is a core component of machine translation, dialogue systems, speech recognition, summarization, and so forth. The existing text generation methods tend to be based on recurrent neural language models (NLMs), which generate sentences from encoding vector. However, most of these models lack explicit structured representation for text generation. In this work, we introduce a new generative model for NLG, called Tree-VAE. First it samples a sentence from the training corpus and then generates a new sentence based on the corresponding parse tree embedding vector. Tree-LSTM is used in collaboration with the Stanford Parser to retrieve sentence construction data, which is then used to train a conditional discretization autoencoder generator based on the embeddings of sentence patterns. The proposed model is extensively evaluated on three different datasets. The experimental results proved that the proposed model can generate substantially more diverse and coherent text than existing baseline methods.

In recent years, the increasingly fierce competition among higher education institutions (HEIs), the finite resources, and the enormous influence of the COVID-19 epidemic on higher education have made it especially important to evaluate the performance of Chinese higher education institutions. This paper utilizes the DEA-BCC and Malmquist index to analyze the efficiency and productivity of 34 Chinese “985 Project” universities in the period 2017–2021. The indicator system includes three inputs and five outputs, contained in Model 1 and Model 2 for comparative analysis. The results demonstrate that the COVID-19 epidemic has had a considerable negative impact on Chinese higher education, and has induced the reduction of technical efficiency and productivity. Setting up online MOOCs is conducive to enhancing the efficiency and productivity of HEIs; in addition, the efficiency mentioned varies noticeably among different university levels, and there is no significant difference in different university types and geographical locations.

In road networks, attribute information carried by road segment nodes, such as weather and points of interest (POI), exhibit strong heterogeneity and often involve one-to-many or many-to-one relationships. However, research on such heterogeneity in traffic prediction is relatively limited. Our research examines how varying the network propagation pattern based on the degree of node-to-node heterogeneity of information affects the model prediction performance. Specifically, at the node level, we use knowledge embedding to generate knowledge vectors that quantify the heterogeneity among the attribute information of a node. At the road network level, we calculate a homogeneity adjacency matrix that captures both the topological structure of the road network and the similarity of node heterogeneity. This adjacency matrix assigns different weights to neighbors based on their homogeneity, guiding the propagation of graph convolutional networks (GCN). Finally, we separate the representation of propagation into self-representation and neighbor representation to extract multi-attribute information, including self, homogeneity, and heterogeneity. Experiments on real datasets demonstrate that the incorporation of our homogeneity adjacency matrix leads to a significant improvement in the accuracy of short-term and long-term prediction compared with previous work on homogeneous and single-dimensional information. Furthermore, our approach maintains its performance advantage over baseline models under different embedding dimensions and parameter settings.

To evaluate the effects of gelatin and starch encapsulated vitamin A on growth performance, immune status and antioxidant capacity in weaned piglets, a total of 96 weaned piglets (body weight = 9.11 ± 0.03 kg, 30-d-old) were randomly allotted to 3 treatments with 4 replications of 8 piglets each. The 3 treatments were control diet (basal diet without addition of vitamin A), gelatin vitamin A diet (basal diet + 13,500 IU/kg gelatin encapsulated vitamin A), and starch vitamin A diet (basal diet + 13,500 IU/kg starch encapsulated vitamin A), respectively. The results showed that piglets fed starch vitamin A diet had significantly higher final body weight and average daily gain compared to those in control and gelatin vitamin A groups (P

Given that traditional anticancer therapies fail to significantly improve the prognoses of triple negative breast cancer (TNBC), new modalities with high efficiency are urgently needed. Herein, by mixing the metal-phenolic network formed by tannic acid (TA), bleomycin (BLM), and Fe3+ with glutathione peroxidase 4 (GPX4) inhibitor (ML210) loaded hollow mesoporous Prussian blue (HMPB) nanocubes, the HMPB/ML210@TA-BLM-Fe3+ (HMTBF) nanocomplex is prepared to favor the ferroptosis/apoptosis synergism in TNBC. During the intracellular degradation, Fe3+ /Fe2+ conversion mediated by TA can initiate the Fenton reaction to drastically upregulate the reactive oxygen species level in cells, subsequently induce the accumulation of lipid peroxidation, and thereby cause ferroptotic cell death; meanwhile, the released ML210 efficiently represses the activity of GPX4 to activate ferroptosis pathway. Besides, the chelation of Fe2+ with BLM leads to in situ BLM toxification at tumor site, then triggers an effective apoptosis to synergize with ferroptosis for tumor therapy. As a result, the superior in vivo antitumor efficacy of HMTBF is corroborated in a 4T1 tumor-bearing mice model regarding tumor growth suppression, indicating that the nanoformulations can serve as efficient ferroptosis and apoptosis inducers for use in combinatorial TNBC therapy.

The controlled synthesis of nanohybrids with magnetic metal oxides (Fe3O4), noble metals Au, and hydroxyapatite (HAp, hydroxyapatite) have received considerable attention for applications in photocatalysis, molecular imaging, drug delivery. Green and multifunctional Fe3O4/HAp/Au composite nanoparticles (NPs) with biocompatibility and magnetism for photocatalyst was developed by a facile method. The overall architecture of nanocomposites is one of the most important factors dictating the physical and chemical properties of nanohybrids with significant advances and various functions. The improvement performance of the Fe3O4/HAp/Au in photocatalysis was evaluated by degradation of organic pollutant (methylene blue, MB) under visible light irradiation. The results showed that the hybrid Fe3O4/HAp/Au NP displayed high photocatalytic activity due to surface plasmon resonance (SPR) of Au, the combination of magnetic NPs and HAps with adsorption, also the mechanism was discussed in detail. The efficient magnetic composite nanoparticles can be separated with target and recover, as well as recycle processes. It proposed novel NPs and methodology to design and fabrication of multifunctional NPs with high quality for application of photocatalysis. Furthermore, it is expected that the safe, green and multifunctional composite NPs will provide an interesting platform not only for photocatalysis, but also for developing molecular imaging probe and drug delivery with target, as well as Raman detection.

Label distribution learning methods effectively address the label ambiguity problem and have achieved great success in image emotion analysis. However, these methods ignore structured and sparse information naturally contained in the annotations of emotions. For example, emotions can be grouped and ordered due to their polarities and degrees. Meanwhile, emotions have the character of intensity and are reflected in different levels of sparse annotations. Motivated by these observations, we present a convolutional neural network based framework called Structured and Sparse annotations for image emotion Distribution Learning (SSDL) to tackle two challenges. In order to utilize structured annotations, the Earth Mover’s Distance is employed to calculate the minimal cost required to transform one distribution to another for ordered emotions and emotion groups. Combined with Kullback-Leibler divergence, we design the loss to penalize the mispredictions according to the dissimilarities of same emotions and different emotions simultaneously. Moreover, in order to handle sparse annotations, sparse regularization based on emotional intensity is adopted. Through combined loss and sparse regularization, SSDL could effectively leverage structured and sparse annotations for predicting emotion distribution. Experiment results demonstrate that our proposed SSDL significantly outperforms the state-of-the-art methods.

In recent years, image style transfer has been greatly improved by using deep learning technology. However, when directly applied to clothing style transfer, the current methods cannot allow the users to self-control the local transfer position of an image, such as separating specific T-shirt or trousers from a figure, and cannot achieve the perfect preservation of clothing shape. Therefore, this paper proposes an interactive image localized style transfer method especially for clothes. We introduce additional image called outline image, which is extracted from content image by interactive algorithm. The interaction consists simply of dragging a rectangle around the desired clothing. Then, we introduce an outline loss function based on distance transform of the outline image, which can achieve the perfect preservation of clothing shape. In order to smooth and denoise the boundary region, total variation regularization is employed. The proposed method constrains that the new style is generated only in the desired clothing part rather than the whole image including background. Therefore, in our new generated images, the original clothing shape can be reserved perfectly. Experiment results show impressive generated clothing images and demonstrate that this is a good approach to design clothes.

As an information carrier with rich semantics, image plays an increasingly important role in real-time monitoring of logistics management. Abnormal objects are typically closely related to the specific region. Detecting abnormal objects in the specific region is conducive to improving the accuracy of detection and analysis, thereby improving the level of logistics management. Motivated by these observations, we design the method called abnormal object detection in a specific region based on Mask R-convolutional neural network: Abnormal Object Detection in Specific Region. In this method, the initial instance segmentation model is obtained by the traditional Mask R-convolutional neural network method, then the region overlap of the specific region is calculated and the overlapping ratio of each instance is determined, and these two parts of information are fused to predict the exceptional object. Finally, the abnormal object is restored and detected in the original image. Experimental results demonstrate that our proposed Abnormal Object Detection in Specific Region can effectively identify abnormal objects in a specific region and significantly outperforms the state-of-the-art methods.

Along with the development of social media on the internet, dialogue systems are becoming more and more intelligent to meet users’ needs for communication, emotion, and social intercourse. Previous studies usually use sequence-to-sequence learning with recurrent neural networks for response generation. However, recurrent-based learning models heavily suffer from the problem of long-distance dependencies in sequences. Moreover, some models neglect crucial information in the dialogue contexts, which leads to uninformative and inflexible responses. To address these issues, we present a bichannel transformer with context encoding (BCTCE) for document-driven conversation. This conversational generator consists of a context encoder, an utterance encoder, and a decoder with attention mechanism. The encoders aim to learn the distributed representation of input texts. The multihop attention mechanism is used in BCTCE to capture the interaction between documents and dialogues. We evaluate the proposed BCTCE by both automatic evaluation and human judgment. The experimental results on the dataset CMU_DoG indicate that the proposed model yields significant improvements over the state-of-the-art baselines on most of the evaluation metrics, and the generated responses of BCTCE are more informative and more relevant to dialogues than baselines.

The authors describe a sensing interface that is capable of selectively adsorbing gold nanopartices (AuNPs). It was applied to electrochemiluminescent (ECL) detection of microRNA (miRNA). The AuNPs are used as signal transduction probes and luminol acts as the ECL reagent. The sensing interface was generated by sequential assembly of Nafion-carbon nanotubes (CNTs) and polyvinylpyrrolidone (PVP) at a glassy carbon electrode (GCE). If only ss-DNA probes and AuNPs are present, the interaction between them leads to the formation of ss-DNA-capped AuNPs. In this case, the capped-AuNP do not assemble at the interface of the modified electrode. As a result, the ECL is weak due to the poor electroconductivity of PVP. Conversely, if ss-DNA probes bind to target miRNA, the AuNPs can’t interact with the DNA/miRNA hybrids formed because of its rigid structure. Hence, aggregated-AuNP are generated. These are concentrated in the sensing interface due to the strong interaction between AuNP and PVP assembled at the modified electrode surface via Au-N chemical bonds. The AuNP concentrated at the surface of the modified GCE electrocatalyze the oxidation of luminol which results in strong ECL. These findings were used to design a quantitative assay for trace levels of miRNA. To confirm the viability of the method, let-7a was used as a model analyte. Under the optimal conditions, the resulting calibration plot is linear in the 40 fM. to 1.0 pM concentration range with a detection limit as low as 20 fM. The method is even capable of discriminating most let-7 miRNA family members and was successfully applied to real sample assay. In our perception, the results demonstrate that this sensing interface represents a cost-effective and sensitive tool for determination of clinically relevant cancer biomarkers.

Recently, automatic visual data understanding from drone platforms becomes highly demanding. To facilitate the study, the Vision Meets Drone Object Detection in Image Challenge is held the second time in conjunction with the 17-th International Conference on Computer Vision (ICCV 2019), focuses on image object detection on drones. Results of 33 object detection algorithms are presented. For each participating detector, a short description is provided in the appendix. Our goal is to advance the state-of-the-art detection algorithms and provide a comprehensive evaluation platform for them. The evaluation protocol of the VisDrone-DET2019 Challenge and the comparison results of all the submitted detectors on the released dataset are publicly available at the website: http: //www.aiskyeye.com/. The results demonstrate that there still remains a large room for improvement for object detection algorithms on drones.

As an information carrier with rich semantics, images contain more sentiment than texts and audios. So, images are increasingly used by people to express their opinions and sentiments in social network. The sentiments of the images are overall and should come from different regions. So, the recognition of the sentiment regions will help to concentrate on important factors the affect the sentiments. Meanwhile, deep learning method for image sentiment classification needs simple and efficient approach for simultaneously carrying out pruning and feature selection whilst optimizing the weights. Motivated by these observations, we design a region-based convolutional neural network using group sparse regularization for image sentiment classification: R-CNNGSR. The method obtains the initial sentiment prediction model through CNN using group sparse regularization to get compact neural network, and then automatically detect the sentiment regions by combining the underlying features and sentimental features. Finally, the whole image and the sentiment region are fused to predict the overall sentiment of the images. Experiment results demonstrate that our proposed R-CNNGSR significantly outperforms the state-of-the-art methods in image sentiment classification.

Video object detection has drawn great attention recently. The Vision Meets Drone Object Detection in Video Challenge 2019 (VisDrone-VID2019) is held to advance the state-of-the-art in video object detection for videos captured by drones. Specifically, there are 13 teams participating the challenge. We also report the results of 6 state-of-the-art detectors on the collected dataset. A short description is provided in the appendix for each participating detector. We present the analysis and discussion of the challenge results. Both the dataset and the challenge results are publicly available at the challenge website: http://www.aiskyeye.com/.

Current state-of-the-art image segmentation methods are mainly based on visible spectrum images. However, it remains challenging under adverse environmental conditions (e.g., nighttime, foggy, snowy). Thermal images can be captured in darkness and deployed in any weather and time. But the research focus on thermal images semantic segmentation remains underexploited since the lack of thermal images segmentation dataset of nighttime driving scene. On the other hand, existing methods do not adequately capture context between related pixels and edge details in thermal images. In this work, we focus on thermal image semantic segmentation of nighttime driving scene. First, we propose a multi-level correction network (MCNet) with a multi-level attention module (MAM) and a multi-level edge enhancement module (MEEM). Specifically, MAM selectively captures the inter-class and intra-class contextual dependencies by a multi-level correction process. In addition, MEEM continuously combines precise context information and edge prior knowledge in each level to correct the final feature representation. Second, we present a new dataset called SCUT-Seg which contains 2010 thermal images recorded from various road scenes with 10 manually annotated semantic region labels. Extensive experiments demonstrate that the proposed method performs favorably against the state-of-the-art methods on SCUT-Seg and the public MFNet dataset. The source code and dataset are available at https://github.com/haitaobiyao/MCNet .

The signal transducer and activator of transcription (STAT) proteins play essential roles in apoptosis, proliferation and survival. However, the role of STATs in intestinal inflammation during weaning is unclear. This study aimed to investigate developmental expression of STATs, nuclear factor-κB (NF-κB) and inflammatory genes in the jejunum of piglets during weaning. Thirty-two piglets were weaned at 21d and sacrificed at 0, 1, 7, or 14d (n=8) after weaning. Villus height and the villus height/crypt depth ratio were decreased, whereas crypt depth was increased in the jejunum at 7 and 14d after weaning. In addition, the mRNA levels of interferon-γ (IFN-γ), inducible nitric oxide synthase (iNOS), IL-6, IL-8, IL-12 and IL-22 were increased in the jejunum at 7 and 14d after weaning, whereas transforming growth factor-β (TGF-β), suppressor of cytokine signaling 3 (SCOS3) and arginase-1 was decreased. Neutrophil infiltration was increased in the mucosa of the jejunum after weaning. Moreover, phosphorylation of IκB-α, NF-κB, AKT and STAT-3 was increased. However, the phosphorylation of STAT-1 (at 7 and 14d) and STAT-6 (at 1 and 7d) was suppressed in the jejunum after weaning. Treatment of porcine jejunal epithelial (IPEC-J2) cells with the STAT inhibitors fludarabine, niclosamide and teriflunomide, which inhibit the phosphorylation of STAT-1, STAT-3 and STAT-6, respectively, weakened the defense capacity of these cells against bacterial infection. In conclusion, weaning caused severe inflammation associated with activation of the NF-κB and STAT-3 pathways and suppression of STAT-1 and STAT-6 in the jejunum of piglets.

We provide an improved algorithm called “a neighborhood expansion tabu search algorithm based on genetic factors” (NETS) to solve traveling salesman problem. The algorithm keeps the traditional tabu algorithm’s neighborhood, ensure the algorithm’s strong climbing ability and go to the local optimization. At the same time, introduce the genetic algorithm’s genetic factor (crossover factor and variation factor) to develop new search space for bounded domain. It can avoid the defects of the alternate search. The results show that this optimization algorithm has improved a lot in the target of “target value”, “convergence” and compared with traditional tabu algorithm and genetic algorithm.

Early nutrition is key to promoting gut growth and education of the immune system. Although iron deficiency anemia has long been recognized as a serious iron disorder, the effects of iron supplementation on gut development are less clear. Therefore, using suckling piglets as the model for iron deficiency, we assessed the impacts of iron supplementation on hematological status, gut development, and immunity improvement. Piglets were parenterally supplied with iron dextran (FeDex, 60 mg Fe/kg) by intramuscular administration on the third day after birth and slaughtered at the age of two days, five days, 10 days, and 20 days. It was expected that iron supplementation with FeDex improved the iron status with higher levels of serum iron, ferritin, transferrin, and iron loading in the liver by regulating the interaction of hepcidin and ferroportin (FPN). FeDex supplementation increased villus length and crypt depth, attenuated the pathological status of the duodenum, and was beneficial to intestinal mucosa. FeDex also influenced the intestinal immune development by stimulating the cytokines&rsquo, production of the intestine and enhancing the phagocytotic capacity of monocytes. Overall, the present study suggested that iron supplementation helped promote the development of the intestine by improving its morphology, which maintains its mucosal integrity and enhances the expression of immuno-associated factors.

Hepcidin, an antimicrobial peptide, is considered to be a key homeostatic regulator of iron metabolism. Tibetan pigs are a Chinese indigenous plateau breed with strong disease resistance. It is therefore of interest to look for porcine hepcidin (pHepc) gene expression in Tibetan pigs in order to clarify the involvement of pHepc in interaction of host defenses and iron homeostasis. In this study, the mRNA expressions of pHepc and the immune response-associated factors were determined by real-time PCR in 30-day-old Tibetan pigs and Landrace pigs were chosen as control. The results showed that the expression of pHepc mRNA in most tissues were higher in Tibetan pigs than those of Landrace pigs, which corresponded to the higher serum iron concentration in Tibetan pigs. In addition, compared to Landrace pigs, higher levels of pro-inflammatory cytokines (TNF-α, IL-1α, IL-1β and IFN-γ), pattern recognition receptors (TLR-2, TLR-4, NOD-1 and NOD-2) and chemokines (MCP-1 and IL-8) were detected in Tibetan pigs, which suggested that Tibetan pigs have a strong immunity. The high expression of pHepc may affect the production of immune factors by regulating iron homeostasis, which may partly explain the strong disease resistance of Tibetan pigs.

Hepcidin was first identified as an antimicrobial peptide and later demonstrated that hepcidin is the long sought hormone to regulate iron homeostasis in mammals. Though its iron regulatory function has been extensively investigated, the studies on its antimicrobial properties are limited. The aim of current study was to evaluate the antibacterial activity of synthetic porcine hepcidin (pHepc) in vitro against pathogen bacteria via radial diffusion, colony forming count, transmission electron microscopy and DNA binding assays. Our results showed that pHepc exerted little bactericidal activity, but possessed bacteriostatic activity by reducing the viable Escherichia coli K88, E. coli ATCC 25922, Staphylococcus aureus ATCC 25923 and Salmonella typhimurium CMCC 50013. pHepc-treated E. coli K88 exhibited longer cells and cytoplasm unevenly distribution, while pHepc-treated S. aureus led to cytoplasm leakage and partly lysis of bacterial cells. Gel retardation assay showed the existence of the binding affinity of pHepc for DNA. In addition, pHepc retained the bacteriostatic activity in a wide range of pH value from 4.0 to 8.0 or in the presence of iron, respectively. Considering the high expression in response to infection and the bacteriostatic activity, pHepc may be an important defense molecule for pig health.

In today's era, the number of today's web text is exploding. The analysis of the text is still a hot topic. The traditional VSM model in the weight statistics and similarity calculation, due to the data latitude is too high, lack of understanding and other issues, will lead to the final clustering inaccurate. In view of this, this paper presents a hybrid model of VSM and LDA for text clustering. Through the collection of text, filtering, application of statistical methods we calculated VSM model and LDA model similarity respectively. The two similarity models are combined by linear addition method, and the mixed similarity is obtained. Then through the K-means algorithm for text clustering and the three models of clustering results we can get the visual effect of clustering. Finally we can judge the merits of the model. The experimental results show that this hybrid model is effective.

Hepcidin, an antimicrobial peptide, was discovered to integrate diverse signals from iron status and an infection threat and orchestrate a series of host-protective responses. Several studies have investigated the antimicrobial role of hepcidin, but the results have been controversial. Here, we aimed to examine the role of hepcidin in bacterial adherence and invasion in vitro . We found that porcine hepcidin could decrease the amount of the extracellular pathogen enterotoxigenic Escherichia coli (ETEC) K88 that adhered to cells because it caused the aggregation of the bacteria. However, addition of hepcidin to macrophages infected with the intracellular pathogen Salmonella enterica serovar Typhimurium enhanced the intracellular growth of the pathogen through the degradation of ferroportin, an iron export protein, and then the sequestration of intracellular iron. Intracellular iron was unavailable by use of the iron chelator deferiprone (DFO), which reduced intracellular bacterial growth. These results demonstrate that hepcidin exhibits different functions in extracellular and intracellular bacterial infections, which suggests that different defense strategies should be taken to prevent bacterial infection.

Class overlapping is thought as one of the toughest problems in data mining because the complex structure of data. The current classification algorithms show little consideration of this problem. So when using this traditional classification algorithms to resolve this problem, classification performance is not good for sample s in overlapping region. To meet this critical challenge, in this paper, we pay a systematic study on the class overlapping problem and propose a new classification algorithm based on NB for class overlapping problem (CANB). CANB uses NB to find class overlapping region and use this region and non-overlapping region in NB classification model learning separately. Experimental results on ben ch mark and real-world data sets demonstrate that CANB can improve the classification performances for class overlapping problem stably and effectively.

Class overlapping and ensemble learning method have long been regarded as one of the toughest yet pervasive problems in data mining. The effect of ensemble learning method and class overlapping are with few researches. To meet this critical challenge, in this paper, we pay a systematic study on the different classifier performance in class overlapping problem and propose a new Ensemble Learning Method based on Naive Bayes for class overlapping problem (ELM-NB). ELM-NB uses NB to find class overlapping region and use this region and non-overlapping region in ensemble learning model. Experimental results on selected data sets prove that ELM-NB can have good classification performance for class overlapping problem.

Fraud in consumer-to-consumer (C2C) e-commerce is becoming more and more serious. The purpose of this study is to develop an effective fraud detection model to assist customers in identifying potential fraud transactions. We use Naive Bayes (NB), decision tree C4.5 and AdaBoost to construct the model for classifying imbalance transaction data, and majority voting is used to combine the model. Several experiments are conducted on Taobao data set to verify the classification performance of the proposed model using four popular performance metrics. The experimental results demonstrate that the model based on NB and AdaBoost&C4.5 can significantly increase the ability to locate potential fraud transactions in C2C e-commerce.

The process of Topic Detection from Short Text Systems (SMS) is to extract distinct topics hidden inside short text collections, such as Twitter, Weibo, and instant messages. With the recent emergence of large volume user generated content collections enabled by online social media, topic detection from SMS becomes a challenging yet promising means for online public opinion analysis. In available literature, many forms and methods of topic detection have been proposed, but obtaining meaningful and coherent data is still difficult to reliably obtain for the extreme sparsity brought by SMS. To this end, we developed a Term-based Consensus Clustering topic detection (TCC) framework to provide an unsupervised methodology for finding distinct topics from within SMS collections. Specifically, we adopt a consensus clustering technique called K-means-based Consensus Clustering to handle SMS clustering, due to its low computational complexity and robust clustering performance. To further enrich the features of the information of the sparse SMS data, we conduct term clustering in the highly dense term space instead of the conventionally targeted sparse document space. To be more specific, we first use a feature space transfer technique to represent short text collections as a pseudo-document matrix, where rows, namely instances, correspond to terms and columns, namely features, correspond to adjacent terms. Basic partitions are generated from the pseudo-document matrix for term clustering and consensus clustering is followed to obtain the final term clustering result. Finally, a document classification process is adopted and a document is assigned to a cluster, where most terms occurred. Extensive experiments on real-world data sets demonstrate that TCC is comparable to several widely used methods in terms of topic detection quality. Particularly, we demonstrate that TCC obtains best clustering performance when observing a large number of the predefined topics across short text collections.