Offensive language detection has received important attention and plays a crucial role in promoting healthy communication on social platforms,as well as promoting the safe deployment of large language models.Training ...Offensive language detection has received important attention and plays a crucial role in promoting healthy communication on social platforms,as well as promoting the safe deployment of large language models.Training data is the basis for developing detectors;however,the available offense-related dataset in Chinese is severely limited in terms of data scale and coverage when compared to English resources.This significantly affects the accuracy of Chinese offensive language detectors in practical applications,especially when dealing with hard cases or out-of-domain samples.To alleviate the limitations posed by available datasets,we introduce AugCOLD(Augmented Chinese Offensive Language Dataset),a large-scale unsupervised dataset containing 1 million samples gathered by data crawling and model generation.Furthermore,we employ a multiteacher distillation framework to enhance detection performance with unsupervised data.That is,we build multiple teachers with publicly accessible datasets and use them to assign soft labels to AugCOLD.The soft labels serve as a bridge for knowledge to be distilled from both AugCOLD and multiteacher to the student network,i.e.,the final offensive detector.We conduct experiments on multiple public test sets and our well-designed hard tests,demonstrating that our proposal can effectively improve the generalization and robustness of the offensive language detector.展开更多
Background:Promoting wound healing is crucial to restore the vital barrier function of injured skin.Growth factor products including epidermal growth factor(EGF),fibroblast growth factor(FGF)and granulocyte-macrophage...Background:Promoting wound healing is crucial to restore the vital barrier function of injured skin.Growth factor products including epidermal growth factor(EGF),fibroblast growth factor(FGF)and granulocyte-macrophage colony stimulating factor(GM-CSF)have been used for decades although no systematic evaluation exists regarding their effectiveness and safety issues in treating acute skin wounds.This has resulted in a lack of guidelines and standards for proper application regimes.Therefore,this systematic review and meta-analysis was performed to critically evaluate the effectiveness and safety of these growth factors on skin acute wounds and provide guidelines for application regimes.Methods:We searched PubMed/Medline(1980-2020),Cochrane Library(1980-2020),Cochrane CENTRAL(from establishment to 2020),ClinicalTrials.gov(from establishment to 2020),Chinese Journal Full-text Database(CNKI,1994-2020),China Biology Medicine disc(CBM,1978-2019),Chinese Scientific Journal Database(VIP,1989-2020)andWanfang Database(WFDATA,1980-2019).Randomized controlled trials(RCTs),quasi-RCTs and controlled clinical trials treating patients with acute skin wounds from various causes and with those available growth factors were included.Results:A total of 7573 papers were identified through database searching;229 papers including 281 studies were kept after final screening.Administering growth factors significantly short-ened the healing time of acute skin wounds,including superficial burn injuries[mean differ-ence(MD)=−3.02;95%confidence interval(CI):−3.31∼−2.74;p<0.00001],deep burn injuries(MD=−5.63;95%CI:−7.10∼−4.17;p<0.00001),traumata and surgical wounds(MD=−4.50;95%CI:−5.55∼−3.44;p<0.00001).Growth factors increased the healing rate of acute skin wounds and decreased scar scores.The incidence of adverse reactions was lower in the growth factor treatment group than in the non-growth factor group.Conclusions:The studied growth factors not only are effective and safe for managing acute skin wounds,but also accelerate their healing with no severe adverse reactions.展开更多
Emotion recognition has been used widely in various applications such as mental health monitoring and emotional management.Usually,emotion recognition is regarded as a text classification task.Emotion recognition is a...Emotion recognition has been used widely in various applications such as mental health monitoring and emotional management.Usually,emotion recognition is regarded as a text classification task.Emotion recognition is a more complex problem,and the relations of emotions expressed in a text are nonnegligible.In this paper,a hierarchical model with label embedding is proposed for contextual emotion recognition.Especially,a hierarchical model is utilized to learn the emotional representation of a given sentence based on its contextual information.To give emotion correlation-based recognition,a label embedding matrix is trained by joint learning,which contributes to the final prediction.Comparison experiments are conducted on Chinese emotional corpus RenCECps,and the experimental results indicate that our approach has a satisfying performance in textual emotion recognition task.展开更多
基金supported by the National Science Foundation for Distinguished Young Scholars(with No.62125604)the NSFC projects(Key project with No.61936010 and regular project with No.61876096)+1 种基金supported by the Guoqiang Institute of Tsinghua University,with Grant No.2019GQG1 and 2020GQG0005sponsored by Tsinghua-Toyota Joint Research Fund.
文摘Offensive language detection has received important attention and plays a crucial role in promoting healthy communication on social platforms,as well as promoting the safe deployment of large language models.Training data is the basis for developing detectors;however,the available offense-related dataset in Chinese is severely limited in terms of data scale and coverage when compared to English resources.This significantly affects the accuracy of Chinese offensive language detectors in practical applications,especially when dealing with hard cases or out-of-domain samples.To alleviate the limitations posed by available datasets,we introduce AugCOLD(Augmented Chinese Offensive Language Dataset),a large-scale unsupervised dataset containing 1 million samples gathered by data crawling and model generation.Furthermore,we employ a multiteacher distillation framework to enhance detection performance with unsupervised data.That is,we build multiple teachers with publicly accessible datasets and use them to assign soft labels to AugCOLD.The soft labels serve as a bridge for knowledge to be distilled from both AugCOLD and multiteacher to the student network,i.e.,the final offensive detector.We conduct experiments on multiple public test sets and our well-designed hard tests,demonstrating that our proposal can effectively improve the generalization and robustness of the offensive language detector.
文摘Background:Promoting wound healing is crucial to restore the vital barrier function of injured skin.Growth factor products including epidermal growth factor(EGF),fibroblast growth factor(FGF)and granulocyte-macrophage colony stimulating factor(GM-CSF)have been used for decades although no systematic evaluation exists regarding their effectiveness and safety issues in treating acute skin wounds.This has resulted in a lack of guidelines and standards for proper application regimes.Therefore,this systematic review and meta-analysis was performed to critically evaluate the effectiveness and safety of these growth factors on skin acute wounds and provide guidelines for application regimes.Methods:We searched PubMed/Medline(1980-2020),Cochrane Library(1980-2020),Cochrane CENTRAL(from establishment to 2020),ClinicalTrials.gov(from establishment to 2020),Chinese Journal Full-text Database(CNKI,1994-2020),China Biology Medicine disc(CBM,1978-2019),Chinese Scientific Journal Database(VIP,1989-2020)andWanfang Database(WFDATA,1980-2019).Randomized controlled trials(RCTs),quasi-RCTs and controlled clinical trials treating patients with acute skin wounds from various causes and with those available growth factors were included.Results:A total of 7573 papers were identified through database searching;229 papers including 281 studies were kept after final screening.Administering growth factors significantly short-ened the healing time of acute skin wounds,including superficial burn injuries[mean differ-ence(MD)=−3.02;95%confidence interval(CI):−3.31∼−2.74;p<0.00001],deep burn injuries(MD=−5.63;95%CI:−7.10∼−4.17;p<0.00001),traumata and surgical wounds(MD=−4.50;95%CI:−5.55∼−3.44;p<0.00001).Growth factors increased the healing rate of acute skin wounds and decreased scar scores.The incidence of adverse reactions was lower in the growth factor treatment group than in the non-growth factor group.Conclusions:The studied growth factors not only are effective and safe for managing acute skin wounds,but also accelerate their healing with no severe adverse reactions.
基金supported in part by the Research Clusters program of Tokushima University under grant no.2003002This research has been partially supported by NSFC-Shenzhen Joint Foundation(Key Project)(Grant no.U1613217).
文摘Emotion recognition has been used widely in various applications such as mental health monitoring and emotional management.Usually,emotion recognition is regarded as a text classification task.Emotion recognition is a more complex problem,and the relations of emotions expressed in a text are nonnegligible.In this paper,a hierarchical model with label embedding is proposed for contextual emotion recognition.Especially,a hierarchical model is utilized to learn the emotional representation of a given sentence based on its contextual information.To give emotion correlation-based recognition,a label embedding matrix is trained by joint learning,which contributes to the final prediction.Comparison experiments are conducted on Chinese emotional corpus RenCECps,and the experimental results indicate that our approach has a satisfying performance in textual emotion recognition task.