Group decision making problems are investigated with uncertain multiplicative linguistic preference relations.An unbalanced multiplicative linguistic label set is introduced,which can be used by the experts to express...Group decision making problems are investigated with uncertain multiplicative linguistic preference relations.An unbalanced multiplicative linguistic label set is introduced,which can be used by the experts to express their linguistic preference information over alternatives.The uncertain linguistic weighted geometric mean operator is utilized to aggregate all the individual uncertain multiplicative linguistic preference relations into a collective one,and then a simple approach is developed to determine the experts＇ weights by utilizing the consensus degrees among the individual uncertain multiplicative linguistic preference relations and the collective uncertain multiplicative linguistic preference relations.Furthermore,a practical interactive procedure for group decision making is proposed based on uncertain multiplicative linguistic preference relations,in which a possibility degree formula and a complementary matrix are used to rank the given alternatives.Finally,the proposed procedure is applied to solve the group decision making problem of a manufacturing company searching the best global supplier for one of its most critical parts used in assembling process.展开更多
细粒度表情识别任务因其包含更丰富真实的人类情感而备受关注.现有面部表情识别算法通过提取局部关键区域等方式学习更优的图像表征.然而,这些方法忽略了图像数据集内在的结构关系,且没有充分利用标签间的语义关联度以及图像和标签间的...细粒度表情识别任务因其包含更丰富真实的人类情感而备受关注.现有面部表情识别算法通过提取局部关键区域等方式学习更优的图像表征.然而,这些方法忽略了图像数据集内在的结构关系,且没有充分利用标签间的语义关联度以及图像和标签间的相关性,导致所学特征带来的性能提升有限.其次,现有细粒度表情识别方法并未有效利用和挖掘粗细粒度的层级关系,因而限制了模型的识别性能.此外,现有细粒度表情识别算法忽略了由于标注主观性和情感复杂性导致的标签歧义性问题,极大影响了模型的识别性能.针对上述问题,本文提出一种基于关系感知和标签消歧的细粒度面部表情识别算法(fine-grained facial expression recognition algorithm based on Relationship-Awareness and Label Disambiguation,RALD).该算法通过构建层级感知的图像特征增强网络,充分挖掘图像之间、层级标签之间以及图像和标签之间的依赖关系,以获得更具辨别性的图像特征.针对标签歧义性问题,算法设计了基于近邻样本的标签分布学习模块,通过整合邻域信息进行标签消歧,进一步提升模型识别性能.在细粒度表情识别数据集FG-Emotions上算法的准确度达到97.34%,在粗粒度表情识别数据集RAF-DB上比现有主流表情分类方法提高了0.80%~4.55%.展开更多
In this paper,we study cross-domain relation extraction.Since new data mapping to feature spaces always differs from the previously seen data due to a domain shif,few-shot relation extraction often perform poorly.To s...In this paper,we study cross-domain relation extraction.Since new data mapping to feature spaces always differs from the previously seen data due to a domain shif,few-shot relation extraction often perform poorly.To solve the problems caused by cross-domain,we propose a method for combining the pure entity,relation labels and adversarial(PERLA).We first use entities and complete sentences for separate encoding to obtain context-independent entity features.Then,we combine relation labels which are useful for relation extraction to mitigate context noise.We combine adversarial to reduce the noise caused by cross-domain.We conducted experiments on the publicly available cross-domain relation extraction dataset Fewrel 2.o[1]o,and the results show that our approach improves accuracy and has better transferability for better adaptation to cross-domain tasks.展开更多
针对玉米育种文本数据中存在重叠三元组、实体表达方式多样等问题,提出一种嵌入词汇信息的BERT-CRF(Bidirectional encoder representations from transformers-conditional random field)玉米育种实体关系联合抽取方法。首先,分析了玉...针对玉米育种文本数据中存在重叠三元组、实体表达方式多样等问题,提出一种嵌入词汇信息的BERT-CRF(Bidirectional encoder representations from transformers-conditional random field)玉米育种实体关系联合抽取方法。首先,分析了玉米育种语料表达特征,采用对实体边界、关系类别和实体位置信息同步标注的策略;其次,构建了嵌入词汇信息的BERT-CRF模型进行训练和预测,自建玉米育种知识词典,通过在BERT中嵌入词汇信息,融合字符特征和词汇特征,增强模型的语义能力,利用CRF模型输出全局最优标签序列,设计了实体关系三元组匹配算法(Entity and relation triple matching algorithm,ERTM),将标签进行匹配和映射来获取三元组;最后,为验证该方法的有效性,在玉米育种数据集上进行实验,结果表明,本文模型精确率、召回率和F1值分别为91.84%、95.84%、93.80%,与现有模型相比性能均有提升。说明该方法能够有效抽取玉米育种领域知识,为构建玉米育种知识图谱及其它下游任务提供数据基础。展开更多
目的:应用Tandem mass tag(TMT)技术联合高内涵筛选(HCS)寻找潜在的胃癌相关基因。方法:利用TMT技术从胃癌及癌旁组织中筛选差异表达蛋白质,应用GO和KEGG数据库进行注释和信号通路富集并结合PubMed数据库结果,筛选出胃癌中未见或较少报...目的:应用Tandem mass tag(TMT)技术联合高内涵筛选(HCS)寻找潜在的胃癌相关基因。方法:利用TMT技术从胃癌及癌旁组织中筛选差异表达蛋白质,应用GO和KEGG数据库进行注释和信号通路富集并结合PubMed数据库结果,筛选出胃癌中未见或较少报道的上调目的基因。通过shRNA技术敲减基因,HCS技术比较敲减对细胞增殖的影响,筛选出肿瘤增殖相关基因并通过检测mRNA含量、克隆形成和CCK8进行验证。结果:通过TMT从胃癌和癌旁组织定量到差异表达蛋白1008个(上调649个,下调359个),25个目的基因敲减后HCS结果显示SF3B4敲减抑制细胞增殖,验证发现SF3B4是潜在的胃癌细胞增殖相关基因。结论:TMT技术结合HCS为肿瘤相关基因的筛选提供新的可能性。本研究提示SF3B4是潜在的胃癌增殖相关基因,为研究胃癌发生机制提供新线索。展开更多
针对目前自动ICD(international classification of diseases)编码任务存在标签空间大、诊断代码分布不均衡与临床文本表征差的问题,提出一种融合Longformer与标签注意力的分层ICD自动编码模型。借助Clinical-Longformer预训练语言模型...针对目前自动ICD(international classification of diseases)编码任务存在标签空间大、诊断代码分布不均衡与临床文本表征差的问题,提出一种融合Longformer与标签注意力的分层ICD自动编码模型。借助Clinical-Longformer预训练语言模型获得融合长文本语境的词向量表征。通过将标签的语义表示与注意力机制相结合,捕捉临床文本中与诊断代码相关的关键特征信息,获取更精准的文本表示。引入分层联合学习机制,建立分层预测层解码输出ICD编码。实验结果表明,该模型的准确率、召回率与F1值均高于现有模型,验证了该方法进行自动ICD编码的有效性,为实施疾病诊断相关分组提供高质量的数据支撑。展开更多
基金supported by the National Natural Science Foundation of China (70571087)the National Science Fund for Distinguished Young Scholars of China (70625005)
文摘Group decision making problems are investigated with uncertain multiplicative linguistic preference relations.An unbalanced multiplicative linguistic label set is introduced,which can be used by the experts to express their linguistic preference information over alternatives.The uncertain linguistic weighted geometric mean operator is utilized to aggregate all the individual uncertain multiplicative linguistic preference relations into a collective one,and then a simple approach is developed to determine the experts＇ weights by utilizing the consensus degrees among the individual uncertain multiplicative linguistic preference relations and the collective uncertain multiplicative linguistic preference relations.Furthermore,a practical interactive procedure for group decision making is proposed based on uncertain multiplicative linguistic preference relations,in which a possibility degree formula and a complementary matrix are used to rank the given alternatives.Finally,the proposed procedure is applied to solve the group decision making problem of a manufacturing company searching the best global supplier for one of its most critical parts used in assembling process.
文摘细粒度表情识别任务因其包含更丰富真实的人类情感而备受关注.现有面部表情识别算法通过提取局部关键区域等方式学习更优的图像表征.然而,这些方法忽略了图像数据集内在的结构关系,且没有充分利用标签间的语义关联度以及图像和标签间的相关性,导致所学特征带来的性能提升有限.其次,现有细粒度表情识别方法并未有效利用和挖掘粗细粒度的层级关系,因而限制了模型的识别性能.此外,现有细粒度表情识别算法忽略了由于标注主观性和情感复杂性导致的标签歧义性问题,极大影响了模型的识别性能.针对上述问题,本文提出一种基于关系感知和标签消歧的细粒度面部表情识别算法(fine-grained facial expression recognition algorithm based on Relationship-Awareness and Label Disambiguation,RALD).该算法通过构建层级感知的图像特征增强网络,充分挖掘图像之间、层级标签之间以及图像和标签之间的依赖关系,以获得更具辨别性的图像特征.针对标签歧义性问题,算法设计了基于近邻样本的标签分布学习模块,通过整合邻域信息进行标签消歧,进一步提升模型识别性能.在细粒度表情识别数据集FG-Emotions上算法的准确度达到97.34%,在粗粒度表情识别数据集RAF-DB上比现有主流表情分类方法提高了0.80%~4.55%.
基金The State Key Program of National Natural Science of China,Grant/Award Number:61533018National Natural Science Foundation of China,Grant/Award Number:61402220+2 种基金The Philosophy and Social Science Foundation of Hunan Province,Grant/Award Number:16YBA323Natural Science Foundation of Hunan Province,Grant/Award Number:2020J4525,2022J30495Scientific Research Fund of Hunan Provincial Education Department,Grant/Award Number:18B279,19A439.
文摘In this paper,we study cross-domain relation extraction.Since new data mapping to feature spaces always differs from the previously seen data due to a domain shif,few-shot relation extraction often perform poorly.To solve the problems caused by cross-domain,we propose a method for combining the pure entity,relation labels and adversarial(PERLA).We first use entities and complete sentences for separate encoding to obtain context-independent entity features.Then,we combine relation labels which are useful for relation extraction to mitigate context noise.We combine adversarial to reduce the noise caused by cross-domain.We conducted experiments on the publicly available cross-domain relation extraction dataset Fewrel 2.o[1]o,and the results show that our approach improves accuracy and has better transferability for better adaptation to cross-domain tasks.
文摘针对玉米育种文本数据中存在重叠三元组、实体表达方式多样等问题,提出一种嵌入词汇信息的BERT-CRF(Bidirectional encoder representations from transformers-conditional random field)玉米育种实体关系联合抽取方法。首先,分析了玉米育种语料表达特征,采用对实体边界、关系类别和实体位置信息同步标注的策略;其次,构建了嵌入词汇信息的BERT-CRF模型进行训练和预测,自建玉米育种知识词典,通过在BERT中嵌入词汇信息,融合字符特征和词汇特征,增强模型的语义能力,利用CRF模型输出全局最优标签序列,设计了实体关系三元组匹配算法(Entity and relation triple matching algorithm,ERTM),将标签进行匹配和映射来获取三元组;最后,为验证该方法的有效性,在玉米育种数据集上进行实验,结果表明,本文模型精确率、召回率和F1值分别为91.84%、95.84%、93.80%,与现有模型相比性能均有提升。说明该方法能够有效抽取玉米育种领域知识,为构建玉米育种知识图谱及其它下游任务提供数据基础。
文摘目的:应用Tandem mass tag(TMT)技术联合高内涵筛选(HCS)寻找潜在的胃癌相关基因。方法:利用TMT技术从胃癌及癌旁组织中筛选差异表达蛋白质,应用GO和KEGG数据库进行注释和信号通路富集并结合PubMed数据库结果,筛选出胃癌中未见或较少报道的上调目的基因。通过shRNA技术敲减基因,HCS技术比较敲减对细胞增殖的影响,筛选出肿瘤增殖相关基因并通过检测mRNA含量、克隆形成和CCK8进行验证。结果:通过TMT从胃癌和癌旁组织定量到差异表达蛋白1008个(上调649个,下调359个),25个目的基因敲减后HCS结果显示SF3B4敲减抑制细胞增殖,验证发现SF3B4是潜在的胃癌细胞增殖相关基因。结论:TMT技术结合HCS为肿瘤相关基因的筛选提供新的可能性。本研究提示SF3B4是潜在的胃癌增殖相关基因,为研究胃癌发生机制提供新线索。
文摘针对目前自动ICD(international classification of diseases)编码任务存在标签空间大、诊断代码分布不均衡与临床文本表征差的问题,提出一种融合Longformer与标签注意力的分层ICD自动编码模型。借助Clinical-Longformer预训练语言模型获得融合长文本语境的词向量表征。通过将标签的语义表示与注意力机制相结合,捕捉临床文本中与诊断代码相关的关键特征信息,获取更精准的文本表示。引入分层联合学习机制,建立分层预测层解码输出ICD编码。实验结果表明,该模型的准确率、召回率与F1值均高于现有模型,验证了该方法进行自动ICD编码的有效性,为实施疾病诊断相关分组提供高质量的数据支撑。
