In the tag recommendation task on academic platforms,existing methods disregard users’customized preferences in favor of extracting tags based just on the content of the articles.Besides,it uses co-occurrence techniq...In the tag recommendation task on academic platforms,existing methods disregard users’customized preferences in favor of extracting tags based just on the content of the articles.Besides,it uses co-occurrence techniques and tries to combine nodes’textual content for modelling.They still do not,however,directly simulate many interactions in network learning.In order to address these issues,we present a novel system that more thoroughly integrates user preferences and citation networks into article labelling recommendations.Specifically,we first employ path similarity to quantify the degree of similarity between user labelling preferences and articles in the citation network.Then,the Commuting Matrix for massive node pair paths is used to improve computational performance.Finally,the two commonalities mentioned above are combined with the interaction paper labels based on the additivity of Poisson distribution.In addition,we also consider solving the model’s parameters by applying variational inference.Experimental results demonstrate that our suggested framework agrees and significantly outperforms the state-of-the-art baseline on two real datasets by efficiently merging the three relational data.Based on the Area Under Curve(AUC)and Mean Average Precision(MAP)analysis,the performance of the suggested task is evaluated,and it is demonstrated to have a greater solving efficiency than current techniques.展开更多
Chinese named entity recognition(CNER)has received widespread attention as an important task of Chinese information extraction.Most previous research has focused on individually studying flat CNER,overlapped CNER,or d...Chinese named entity recognition(CNER)has received widespread attention as an important task of Chinese information extraction.Most previous research has focused on individually studying flat CNER,overlapped CNER,or discontinuous CNER.However,a unified CNER is often needed in real-world scenarios.Recent studies have shown that grid tagging-based methods based on character-pair relationship classification hold great potential for achieving unified NER.Nevertheless,how to enrich Chinese character-pair grid representations and capture deeper dependencies between character pairs to improve entity recognition performance remains an unresolved challenge.In this study,we enhance the character-pair grid representation by incorporating both local and global information.Significantly,we introduce a new approach by considering the character-pair grid representation matrix as a specialized image,converting the classification of character-pair relationships into a pixel-level semantic segmentation task.We devise a U-shaped network to extract multi-scale and deeper semantic information from the grid image,allowing for a more comprehensive understanding of associative features between character pairs.This approach leads to improved accuracy in predicting their relationships,ultimately enhancing entity recognition performance.We conducted experiments on two public CNER datasets in the biomedical domain,namely CMeEE-V2 and Diakg.The results demonstrate the effectiveness of our approach,which achieves F1-score improvements of 7.29 percentage points and 1.64 percentage points compared to the current state-of-the-art(SOTA)models,respectively.展开更多
This study was to explore the functional mechanism of rare earth regulating soybean leaves and the characteristics and functions of differentially expressed proteins under the regulation of rare earth. In this study, ...This study was to explore the functional mechanism of rare earth regulating soybean leaves and the characteristics and functions of differentially expressed proteins under the regulation of rare earth. In this study, Dongnong 42 was used as material, and 30 mg·L^(-1) CeCl_(3) solution was sprayed on soybean leaves at the seedling stage. Tandem mass tag(TMT) quantitative proteomics technique and bioinformatics analysis were used to identify soybean leaf proteins. A total of 8 510 proteins were identified, and 127 differentially expressed proteins(DEPs) in response to rare earth cerium regulation were identified, among which 64 were upregulated and 63 were down-regulated. The gene ontology(GO) annotation indicated that DEPs were mainly involved in metabolic process, cellular process, response to stimulus, biological regulation, and response to a stimulus;DEPs in cell module categories were mainly involved in cells, cell part, organelle, membrane, membrane part, organelle par, and protein-containing complex;DEPs in molecular functional categories were mainly involved in catalytic activity, binding and antioxidant activity. Kyoto encyclopedia of genes and genomes(KEGG) pathway significantly enriched starch and sucrose metabolism, glycolysis/gluconeogenesis, galactose metabolism, pentose phosphate pathway, and MAPK signaling pathway-plant. These DEPs were mainly involved in photosynthesis, glucose metabolism and stress response. Forty-six differential protein interaction networks were identified by protein interaction network analysis. This experiment provided a reference for studies of the mechanism of rare earth cerium regulating soybean leaf function from the proteomic perspective.展开更多
BACKGROUND As a well-known fact to the public,gestational diabetes mellitus(GDM)could bring serious risks for both pregnant women and infants.During this important investigation into the linkage between GDM patients a...BACKGROUND As a well-known fact to the public,gestational diabetes mellitus(GDM)could bring serious risks for both pregnant women and infants.During this important investigation into the linkage between GDM patients and their altered expression in the serum,proteomics techniques were deployed to detect the differentially expressed proteins(DEPs)of in the serum of GDM patients to further explore its pathogenesis,and find out possible biomarkers to forecast GDM occurrence.METHODS Subjects were divided into GDM and normal control groups according to the IADPSG diagnostic criteria.Serum samples were randomly selected from four cases in each group at 24-28 wk of gestation,and the blood samples were identified by applying iTRAQ technology combined with liquid chromatography-tandem mass spectrometry.Key proteins and signaling pathways associated with GDM were identified by bioinformatics analysis,and the expression of key proteins in serum from 12 wk to 16 wk of gestation was further verified using enzyme-linked immunosorbent assay (ELISA).RESULTS Forty-seven proteins were significantly differentially expressed by analyzing the serum samples between the GDMgravidas as well as the healthy ones. Among them, 31 proteins were found to be upregulated notably and the rest16 proteins were downregulated remarkably. Bioinformatic data report revealed abnormal expression of proteinsassociated with lipid metabolism, coagulation cascade activation, complement system and inflammatory responsein the GDM group. ELISA results showed that the contents of RBP4, as well as ANGPTL8, increased in the serumof GDM gravidas compared with the healthy ones, and this change was found to initiate from 12 wk to 16 wk ofgestation.CONCLUSION GDM symptoms may involve abnormalities in lipid metabolism, coagulation cascade activation, complementsystem and inflammatory response. RBP4 and ANGPTL8 are expected to be early predictors of GDM.展开更多
This review article reports the recent progress in the development of a new group of molecule-based flow diagnostic techniques, which include molecular tag- ging velocimetry (MTV) and molecular tagging thermometry ...This review article reports the recent progress in the development of a new group of molecule-based flow diagnostic techniques, which include molecular tag- ging velocimetry (MTV) and molecular tagging thermometry (MTT), for both qualitative flow visualization of thermally induced flow structures and quantitative whole-field mea- surements of flow velocity and temperature distributions. The MTV and MTT techniques can also be easily combined to result in a so-called molecular tagging velocimetry and ther- mometry (MTV&T) technique, which is capble of achieving simultaneous measurements of flow velocity and temperature distribution in fluid flows. Instead of using tiny particles, the molecular tagging techniques (MTV, MTT, and MTV&T) use phosphorescent molecules, which can be turned into long-lasting glowing marks upon excitation by photons of appropriate wavelength, as the tracers for the flow veloc- ity and temperature measurements. The unique attraction and implementation of the molecular tagging techniques are demonstrated by three application examples, which include: (1) to quantify the unsteady heat transfer process from a heated cylinder to the surrounding fluid flow in order to exam- ine the thermal effects on the wake instabilities behind the heated cylinder operating in mixed and forced heat convec- tion regimes, (2) to reveal the time evolution of unsteady heat transfer and phase changing process inside micro-sized, icing water droplets in order to elucidate the underlying physics pertinent to aircraft icing phenomena, and (3) to achievesimultaneous droplet size, velocity and temperature measure- ments of "in-flight" droplets to characterize the dynamic and thermodynamic behaviors of flying droplets in spray flows.展开更多
Purpose: Myocardial fibrosis causes cardiac dysfunction, arrhythmias, and sudden death. Tagging imaging on cardiovascular MR can measure the intra-myocardial motion from the dynamic deformation of lines superimposed o...Purpose: Myocardial fibrosis causes cardiac dysfunction, arrhythmias, and sudden death. Tagging imaging on cardiovascular MR can measure the intra-myocardial motion from the dynamic deformation of lines superimposed on the myocardium. The purpose of this study was to evaluate the detectability of myocardial fibrosis using tagging imaging and to compare this with conventional cine imaging. Materials and Methods: We reviewed 4 normal control (NML) subjects, 4 patients with myocarditis (MYO), and 4 patients with old myocardial infarction (ICM). We measured circumferential strain (Ecc) from tagging imaging, and regional wall thickening (rWT) from cine imaging. Fibrosis was determined from a late gadolinium enhancement (LGE) image. We evaluate diagnostic performance by comparing values of the area under curve (AUC) using ROC analysis. Results: Mean values of Ecc and rWT decreased in the area of LGE both in MYO and ICM patients. AUC values of Ecc and rWT in all subjects were 0.98 and 0.84, respectively (p < 0.0001). These values in MYO patients were 0.95 and 0.72 (p = 0.007), respectively, and 0.99 and 0.75, respectively, in ICM patients (p = 0.0008). Conclusions: Both Ecc and rWT decreased in the area with fibrosis in the patients with MYO and ICM. Tagging imaging showed better detectability of myocardial fibrosis than did cine imaging.展开更多
The paper proposes a unified framework to combine the advantages of the fast one-at-a-time approach and the high-performance all-at-once approach to perform Chinese Word Segmentation(CWS) and Part-of-Speech(PoS) taggi...The paper proposes a unified framework to combine the advantages of the fast one-at-a-time approach and the high-performance all-at-once approach to perform Chinese Word Segmentation(CWS) and Part-of-Speech(PoS) tagging.In this framework,the input of the PoS tagger is a candidate set of several CWS results provided by the CWS model.The widely used one-at-a-time approach and all-at-once approach are two extreme cases of the proposed candidate-based approaches.Experiments on Penn Chinese Treebank 5 and Tsinghua Chinese Treebank show that the generalized candidate-based approach outperforms one-at-a-time approach and even the all-at-once approach.The candidate-based approach is also faster than the time-consuming all-at-once approach.The authors compare three different methods based on sentence,words and character-intervals to generate the candidate set.It turns out that the word-based method has the best performance.展开更多
In order to overcome defects of the classical hidden Markov model (HMM), Markov family model (MFM), a new statistical model was proposed. Markov family model was applied to speech recognition and natural language proc...In order to overcome defects of the classical hidden Markov model (HMM), Markov family model (MFM), a new statistical model was proposed. Markov family model was applied to speech recognition and natural language processing. The speaker independently continuous speech recognition experiments and the part-of-speech tagging experiments show that Markov family model has higher performance than hidden Markov model. The precision is enhanced from 94.642% to 96.214% in the part-of-speech tagging experiments, and the work rate is reduced by 11.9% in the speech recognition experiments with respect to HMM baseline system.展开更多
A method of part-of-speech tagging of English text based on closed-words, wold-form and rules, its abstract model and formal description of its realizing procedure are presented. Finally, an experimental example is gi...A method of part-of-speech tagging of English text based on closed-words, wold-form and rules, its abstract model and formal description of its realizing procedure are presented. Finally, an experimental example is givento illustrate the application of this method.展开更多
Hidden Markov Model(HMM) is a main solution to ambiguities in Chinese segmentation and POS (part-of-speech) tagging. While most previous works for HMM-based Chinese segmentation and POS tagging consult POS information...Hidden Markov Model(HMM) is a main solution to ambiguities in Chinese segmentation and POS (part-of-speech) tagging. While most previous works for HMM-based Chinese segmentation and POS tagging consult POS information in contexts, they do not utilize lexical information which is crucial for resolving certain morphological ambiguity. This paper proposes a method which incorporates lexical information and wider context information into HMM. Model induction and related smoothing technique are presented in detail. Experiments indicate that this technique improves the segmentation and tagging accuracy by nearly 1%.展开更多
Previous studies have shown that there is potential semantic dependency between part-of-speech and semantic roles.At the same time,the predicate-argument structure in a sentence is important information for semantic r...Previous studies have shown that there is potential semantic dependency between part-of-speech and semantic roles.At the same time,the predicate-argument structure in a sentence is important information for semantic role labeling task.In this work,we introduce the auxiliary deep neural network model,which models semantic dependency between part-of-speech and semantic roles and incorporates the information of predicate-argument into semantic role labeling.Based on the framework of joint learning,part-of-speech tagging is used as an auxiliary task to improve the result of the semantic role labeling.In addition,we introduce the argument recognition layer in the training process of the main task-semantic role labeling,so the argument-related structural information selected by the predicate through the attention mechanism is used to assist the main task.Because the model makes full use of the semantic dependency between part-of-speech and semantic roles and the structural information of predicate-argument,our model achieved the F1 value of 89.0%on the WSJ test set of CoNLL2005,which is superior to existing state-of-the-art model about 0.8%.展开更多
Proteomics is a powerful tool that can be used to elucidate the underlying mechanisms of diseases and identify new biomarkers.Therefore,it may also be helpful for understanding the detailed pathological mechanism of t...Proteomics is a powerful tool that can be used to elucidate the underlying mechanisms of diseases and identify new biomarkers.Therefore,it may also be helpful for understanding the detailed pathological mechanism of traumatic brain injury(TBI).In this study,we performed Tandem Mass Tag-based quantitative analysis of cortical proteome profiles in a mouse model of TBI.Our results showed that there were 302 differentially expressed proteins in TBI mice compared with normal mice 7 days after injury.Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway analyses showed that these differentially expressed proteins were predominantly involved in inflammatory responses,including complement and coagulation cascades,as well as chemokine signaling pathways.Subsequent transcription factor analysis revealed that the inflammation-related transcription factors NF-κB1,RelA,IRF1,STAT1,and Spi1 play pivotal roles in the secondary injury that occurs after TBI,which further corroborates the functional enrichment for inflammatory factors.Our results suggest that inflammation-related proteins and inflammatory responses are promising targets for the treatment of TBI.展开更多
基金supported by the National Natural Science Foundation of China(No.62271274).
文摘In the tag recommendation task on academic platforms,existing methods disregard users’customized preferences in favor of extracting tags based just on the content of the articles.Besides,it uses co-occurrence techniques and tries to combine nodes’textual content for modelling.They still do not,however,directly simulate many interactions in network learning.In order to address these issues,we present a novel system that more thoroughly integrates user preferences and citation networks into article labelling recommendations.Specifically,we first employ path similarity to quantify the degree of similarity between user labelling preferences and articles in the citation network.Then,the Commuting Matrix for massive node pair paths is used to improve computational performance.Finally,the two commonalities mentioned above are combined with the interaction paper labels based on the additivity of Poisson distribution.In addition,we also consider solving the model’s parameters by applying variational inference.Experimental results demonstrate that our suggested framework agrees and significantly outperforms the state-of-the-art baseline on two real datasets by efficiently merging the three relational data.Based on the Area Under Curve(AUC)and Mean Average Precision(MAP)analysis,the performance of the suggested task is evaluated,and it is demonstrated to have a greater solving efficiency than current techniques.
基金supported by Yunnan Provincial Major Science and Technology Special Plan Projects(Grant Nos.202202AD080003,202202AE090008,202202AD080004,202302AD080003)National Natural Science Foundation of China(Grant Nos.U21B2027,62266027,62266028,62266025)Yunnan Province Young and Middle-Aged Academic and Technical Leaders Reserve Talent Program(Grant No.202305AC160063).
文摘Chinese named entity recognition(CNER)has received widespread attention as an important task of Chinese information extraction.Most previous research has focused on individually studying flat CNER,overlapped CNER,or discontinuous CNER.However,a unified CNER is often needed in real-world scenarios.Recent studies have shown that grid tagging-based methods based on character-pair relationship classification hold great potential for achieving unified NER.Nevertheless,how to enrich Chinese character-pair grid representations and capture deeper dependencies between character pairs to improve entity recognition performance remains an unresolved challenge.In this study,we enhance the character-pair grid representation by incorporating both local and global information.Significantly,we introduce a new approach by considering the character-pair grid representation matrix as a specialized image,converting the classification of character-pair relationships into a pixel-level semantic segmentation task.We devise a U-shaped network to extract multi-scale and deeper semantic information from the grid image,allowing for a more comprehensive understanding of associative features between character pairs.This approach leads to improved accuracy in predicting their relationships,ultimately enhancing entity recognition performance.We conducted experiments on two public CNER datasets in the biomedical domain,namely CMeEE-V2 and Diakg.The results demonstrate the effectiveness of our approach,which achieves F1-score improvements of 7.29 percentage points and 1.64 percentage points compared to the current state-of-the-art(SOTA)models,respectively.
基金Supported by the National Natural Science Foundation of China(31471440)。
文摘This study was to explore the functional mechanism of rare earth regulating soybean leaves and the characteristics and functions of differentially expressed proteins under the regulation of rare earth. In this study, Dongnong 42 was used as material, and 30 mg·L^(-1) CeCl_(3) solution was sprayed on soybean leaves at the seedling stage. Tandem mass tag(TMT) quantitative proteomics technique and bioinformatics analysis were used to identify soybean leaf proteins. A total of 8 510 proteins were identified, and 127 differentially expressed proteins(DEPs) in response to rare earth cerium regulation were identified, among which 64 were upregulated and 63 were down-regulated. The gene ontology(GO) annotation indicated that DEPs were mainly involved in metabolic process, cellular process, response to stimulus, biological regulation, and response to a stimulus;DEPs in cell module categories were mainly involved in cells, cell part, organelle, membrane, membrane part, organelle par, and protein-containing complex;DEPs in molecular functional categories were mainly involved in catalytic activity, binding and antioxidant activity. Kyoto encyclopedia of genes and genomes(KEGG) pathway significantly enriched starch and sucrose metabolism, glycolysis/gluconeogenesis, galactose metabolism, pentose phosphate pathway, and MAPK signaling pathway-plant. These DEPs were mainly involved in photosynthesis, glucose metabolism and stress response. Forty-six differential protein interaction networks were identified by protein interaction network analysis. This experiment provided a reference for studies of the mechanism of rare earth cerium regulating soybean leaf function from the proteomic perspective.
基金This study was reviewed and approved by the Maternal and child health hospital of Hubei Province(Approval No.20201025).
文摘BACKGROUND As a well-known fact to the public,gestational diabetes mellitus(GDM)could bring serious risks for both pregnant women and infants.During this important investigation into the linkage between GDM patients and their altered expression in the serum,proteomics techniques were deployed to detect the differentially expressed proteins(DEPs)of in the serum of GDM patients to further explore its pathogenesis,and find out possible biomarkers to forecast GDM occurrence.METHODS Subjects were divided into GDM and normal control groups according to the IADPSG diagnostic criteria.Serum samples were randomly selected from four cases in each group at 24-28 wk of gestation,and the blood samples were identified by applying iTRAQ technology combined with liquid chromatography-tandem mass spectrometry.Key proteins and signaling pathways associated with GDM were identified by bioinformatics analysis,and the expression of key proteins in serum from 12 wk to 16 wk of gestation was further verified using enzyme-linked immunosorbent assay (ELISA).RESULTS Forty-seven proteins were significantly differentially expressed by analyzing the serum samples between the GDMgravidas as well as the healthy ones. Among them, 31 proteins were found to be upregulated notably and the rest16 proteins were downregulated remarkably. Bioinformatic data report revealed abnormal expression of proteinsassociated with lipid metabolism, coagulation cascade activation, complement system and inflammatory responsein the GDM group. ELISA results showed that the contents of RBP4, as well as ANGPTL8, increased in the serumof GDM gravidas compared with the healthy ones, and this change was found to initiate from 12 wk to 16 wk ofgestation.CONCLUSION GDM symptoms may involve abnormalities in lipid metabolism, coagulation cascade activation, complementsystem and inflammatory response. RBP4 and ANGPTL8 are expected to be early predictors of GDM.
基金supported by the National Aeronauticaland Space Administration(NASA)(Grant NNX12AC21A)The support of the National Science Foundation(NSF)under award numbers of CBET-1064196,IIA-1064235 and CBET-1435590
文摘This review article reports the recent progress in the development of a new group of molecule-based flow diagnostic techniques, which include molecular tag- ging velocimetry (MTV) and molecular tagging thermometry (MTT), for both qualitative flow visualization of thermally induced flow structures and quantitative whole-field mea- surements of flow velocity and temperature distributions. The MTV and MTT techniques can also be easily combined to result in a so-called molecular tagging velocimetry and ther- mometry (MTV&T) technique, which is capble of achieving simultaneous measurements of flow velocity and temperature distribution in fluid flows. Instead of using tiny particles, the molecular tagging techniques (MTV, MTT, and MTV&T) use phosphorescent molecules, which can be turned into long-lasting glowing marks upon excitation by photons of appropriate wavelength, as the tracers for the flow veloc- ity and temperature measurements. The unique attraction and implementation of the molecular tagging techniques are demonstrated by three application examples, which include: (1) to quantify the unsteady heat transfer process from a heated cylinder to the surrounding fluid flow in order to exam- ine the thermal effects on the wake instabilities behind the heated cylinder operating in mixed and forced heat convec- tion regimes, (2) to reveal the time evolution of unsteady heat transfer and phase changing process inside micro-sized, icing water droplets in order to elucidate the underlying physics pertinent to aircraft icing phenomena, and (3) to achievesimultaneous droplet size, velocity and temperature measure- ments of "in-flight" droplets to characterize the dynamic and thermodynamic behaviors of flying droplets in spray flows.
文摘Purpose: Myocardial fibrosis causes cardiac dysfunction, arrhythmias, and sudden death. Tagging imaging on cardiovascular MR can measure the intra-myocardial motion from the dynamic deformation of lines superimposed on the myocardium. The purpose of this study was to evaluate the detectability of myocardial fibrosis using tagging imaging and to compare this with conventional cine imaging. Materials and Methods: We reviewed 4 normal control (NML) subjects, 4 patients with myocarditis (MYO), and 4 patients with old myocardial infarction (ICM). We measured circumferential strain (Ecc) from tagging imaging, and regional wall thickening (rWT) from cine imaging. Fibrosis was determined from a late gadolinium enhancement (LGE) image. We evaluate diagnostic performance by comparing values of the area under curve (AUC) using ROC analysis. Results: Mean values of Ecc and rWT decreased in the area of LGE both in MYO and ICM patients. AUC values of Ecc and rWT in all subjects were 0.98 and 0.84, respectively (p < 0.0001). These values in MYO patients were 0.95 and 0.72 (p = 0.007), respectively, and 0.99 and 0.75, respectively, in ICM patients (p = 0.0008). Conclusions: Both Ecc and rWT decreased in the area with fibrosis in the patients with MYO and ICM. Tagging imaging showed better detectability of myocardial fibrosis than did cine imaging.
基金supported by the National Natural Science Foundation of China under GrantNo.60873174
文摘The paper proposes a unified framework to combine the advantages of the fast one-at-a-time approach and the high-performance all-at-once approach to perform Chinese Word Segmentation(CWS) and Part-of-Speech(PoS) tagging.In this framework,the input of the PoS tagger is a candidate set of several CWS results provided by the CWS model.The widely used one-at-a-time approach and all-at-once approach are two extreme cases of the proposed candidate-based approaches.Experiments on Penn Chinese Treebank 5 and Tsinghua Chinese Treebank show that the generalized candidate-based approach outperforms one-at-a-time approach and even the all-at-once approach.The candidate-based approach is also faster than the time-consuming all-at-once approach.The authors compare three different methods based on sentence,words and character-intervals to generate the candidate set.It turns out that the word-based method has the best performance.
基金Project(60763001)supported by the National Natural Science Foundation of ChinaProjects(2009GZS0027,2010GZS0072)supported by the Natural Science Foundation of Jiangxi Province,China
文摘In order to overcome defects of the classical hidden Markov model (HMM), Markov family model (MFM), a new statistical model was proposed. Markov family model was applied to speech recognition and natural language processing. The speaker independently continuous speech recognition experiments and the part-of-speech tagging experiments show that Markov family model has higher performance than hidden Markov model. The precision is enhanced from 94.642% to 96.214% in the part-of-speech tagging experiments, and the work rate is reduced by 11.9% in the speech recognition experiments with respect to HMM baseline system.
文摘A method of part-of-speech tagging of English text based on closed-words, wold-form and rules, its abstract model and formal description of its realizing procedure are presented. Finally, an experimental example is givento illustrate the application of this method.
基金国家高技术研究发展计划(863计划),the National Natural Science Foundation of China
文摘Hidden Markov Model(HMM) is a main solution to ambiguities in Chinese segmentation and POS (part-of-speech) tagging. While most previous works for HMM-based Chinese segmentation and POS tagging consult POS information in contexts, they do not utilize lexical information which is crucial for resolving certain morphological ambiguity. This paper proposes a method which incorporates lexical information and wider context information into HMM. Model induction and related smoothing technique are presented in detail. Experiments indicate that this technique improves the segmentation and tagging accuracy by nearly 1%.
基金The work of this article is supported by Key Scientific Research Projects of Colleges and Universities in Henan Province(Grant No.20A520007)National Natural Science Foundation of China(Grant No.61402149).
文摘Previous studies have shown that there is potential semantic dependency between part-of-speech and semantic roles.At the same time,the predicate-argument structure in a sentence is important information for semantic role labeling task.In this work,we introduce the auxiliary deep neural network model,which models semantic dependency between part-of-speech and semantic roles and incorporates the information of predicate-argument into semantic role labeling.Based on the framework of joint learning,part-of-speech tagging is used as an auxiliary task to improve the result of the semantic role labeling.In addition,we introduce the argument recognition layer in the training process of the main task-semantic role labeling,so the argument-related structural information selected by the predicate through the attention mechanism is used to assist the main task.Because the model makes full use of the semantic dependency between part-of-speech and semantic roles and the structural information of predicate-argument,our model achieved the F1 value of 89.0%on the WSJ test set of CoNLL2005,which is superior to existing state-of-the-art model about 0.8%.
基金supported by the National Natural Science Foundation of China,No. 81771327a grant for the Platform Construction of Basic Research and Clinical Translation of Nervous System Injury,China,No. PXM2020_026280_000002 (both to BYL)
文摘Proteomics is a powerful tool that can be used to elucidate the underlying mechanisms of diseases and identify new biomarkers.Therefore,it may also be helpful for understanding the detailed pathological mechanism of traumatic brain injury(TBI).In this study,we performed Tandem Mass Tag-based quantitative analysis of cortical proteome profiles in a mouse model of TBI.Our results showed that there were 302 differentially expressed proteins in TBI mice compared with normal mice 7 days after injury.Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway analyses showed that these differentially expressed proteins were predominantly involved in inflammatory responses,including complement and coagulation cascades,as well as chemokine signaling pathways.Subsequent transcription factor analysis revealed that the inflammation-related transcription factors NF-κB1,RelA,IRF1,STAT1,and Spi1 play pivotal roles in the secondary injury that occurs after TBI,which further corroborates the functional enrichment for inflammatory factors.Our results suggest that inflammation-related proteins and inflammatory responses are promising targets for the treatment of TBI.