随着互联网和面向服务技术的发展,一种新型的Web应用——Mashup服务,开始在互联网上流行并快速增长.如何在众多Mashup服务中找到高质量的服务,已经成为一个大家关注的热点问题.寻找功能相似的服务并进行聚类,能有效提升服务发现的精度...随着互联网和面向服务技术的发展,一种新型的Web应用——Mashup服务,开始在互联网上流行并快速增长.如何在众多Mashup服务中找到高质量的服务,已经成为一个大家关注的热点问题.寻找功能相似的服务并进行聚类,能有效提升服务发现的精度与效率.目前国内外主流方法为挖掘Mashup服务中隐含的功能信息,进一步采用特定聚类算法如K-means等进行聚类.然而Mashup服务文档通常为短文本,基于传统的挖掘算法如LDA无法有效处理短文本,导致聚类效果并不理想.针对这一问题,提出一种基于非负矩阵分解的TWE-NMF(nonnegative matrix factorization combining tags and word embedding)模型对Mashup服务进行主题建模.所提方法首先对Mashup服务规范化处理,其次采用一种基于改进的Gibbs采样的狄利克雷过程混合模型,自动估算主题的数量,随后将词嵌入和服务标签等信息与非负矩阵分解相结合,求解Mashup服务主题特征,并通过谱聚类算法将服务聚类.最后,对所提方法的性能进行了综合评价,实验结果表明,与现有的服务聚类方法相比,所提方法在准确率、召回率、F-measure、纯度和熵等评价指标方面都有显著提高.展开更多
This study was conducted to analyze the variation of soil multifunctionality(SMF)along elevation and the driving factors in the Altun Shan.Soil samples(0–10 cm)were collected from 15 sites(H01 to H15)at every 200 m e...This study was conducted to analyze the variation of soil multifunctionality(SMF)along elevation and the driving factors in the Altun Shan.Soil samples(0–10 cm)were collected from 15 sites(H01 to H15)at every 200 m elevation interval,covering a total range from 900 m to 3500 m above mean sea level.We investigated climate factors(mean annual temperature,MAT;mean annual precipitation,MAP),soil environment(soil water content,electrical conductance,and pH),vegetation factors,and elevation to determine which of them are the main driving factors of the spatial variability of SMF in the Altun Shan.We explored the best-fit model of SMF along the changes in elevation using a structural equation model,performed variance partitioning analysis(VPA)on SMF with the“varpart”function to explain the relative contribution of various environmental factors to SMF changes,and used a random forest model for relative importance analysis.The results showed that SMF in the Altun Shan significantly increased with elevation in a linear trend.The main driver of changes in SMF was found to be MAP.Although the rise in elevation did not have a significant direct effect on changes in SMF,it could indirectly affect SMF by significantly influencing MAP,p H,MAT,and normalized difference vegetation index(NDVI).When considering climate,soil environment,and vegetation factors together,they explained 76%of the variation in SMF.The largest contribution to the variation in SMF was attributed to the independent effect of climate(0.31)and its interactive effect with soil(0.30).The relative importance of MAP on SMF changes was found to be the greatest.It is indicated that changes in SMF are caused by the combined effect of multiple environmental conditions.These findings are essential for understanding the spatial variability and drivers of SMF in dryland mountain ecosystems,especially concerning the function of mountain ecosystems in the context of global climatic changes.展开更多
随着互联网上Mashup服务数量及种类的急剧增长,如何从这些海量的服务集合中快速、精准地发现满足用户需求的Mashup服务,成为一个具有挑战性的问题.针对这一问题,本文提出一种融合功能语义关联计算与密度峰值检测的Mashup服务聚类方法,...随着互联网上Mashup服务数量及种类的急剧增长,如何从这些海量的服务集合中快速、精准地发现满足用户需求的Mashup服务,成为一个具有挑战性的问题.针对这一问题,本文提出一种融合功能语义关联计算与密度峰值检测的Mashup服务聚类方法,用于缩小服务的搜索空间,提升服务发现的精度与效率.首先,该方法对Mashup服务进行元信息提取和描述文本内容整理,并根据Web API组合的标签对相应Mashup服务标签进行扩充.然后,用基于功能语义关联计算方法(Functional Semantic Association Calculation Method,FSAC)提取出各服务描述的功能名词集合,并通过功能名词的语义权重来构造Mashup语义特征向量.最后,通过基于密度信息的聚类中心检测方法(Clustering Center Detection Method based on Density Information,CCD-DI)检测出最为合适的K个Mashup语义特征向量作为K-means算法的初始中心,进行聚类划分.基于ProgrammableWeb的真实数据实验表明,本文所提聚类方法在纯度、精准率、召回率、熵等指标上均有良好表现.展开更多
文摘随着互联网和面向服务技术的发展,一种新型的Web应用——Mashup服务,开始在互联网上流行并快速增长.如何在众多Mashup服务中找到高质量的服务,已经成为一个大家关注的热点问题.寻找功能相似的服务并进行聚类,能有效提升服务发现的精度与效率.目前国内外主流方法为挖掘Mashup服务中隐含的功能信息,进一步采用特定聚类算法如K-means等进行聚类.然而Mashup服务文档通常为短文本,基于传统的挖掘算法如LDA无法有效处理短文本,导致聚类效果并不理想.针对这一问题,提出一种基于非负矩阵分解的TWE-NMF(nonnegative matrix factorization combining tags and word embedding)模型对Mashup服务进行主题建模.所提方法首先对Mashup服务规范化处理,其次采用一种基于改进的Gibbs采样的狄利克雷过程混合模型,自动估算主题的数量,随后将词嵌入和服务标签等信息与非负矩阵分解相结合,求解Mashup服务主题特征,并通过谱聚类算法将服务聚类.最后,对所提方法的性能进行了综合评价,实验结果表明,与现有的服务聚类方法相比,所提方法在准确率、召回率、F-measure、纯度和熵等评价指标方面都有显著提高.
基金the Tianshan Talent Plan(2022TSYCCX0001)Natural Science Foundation of Xinjiang(2022D01D083)+1 种基金the National Natural Science Foundation of China(U2003214 and 41977099)the Ecological Processes and Biological Adaptation team for financial and experimental instrumentation help。
文摘This study was conducted to analyze the variation of soil multifunctionality(SMF)along elevation and the driving factors in the Altun Shan.Soil samples(0–10 cm)were collected from 15 sites(H01 to H15)at every 200 m elevation interval,covering a total range from 900 m to 3500 m above mean sea level.We investigated climate factors(mean annual temperature,MAT;mean annual precipitation,MAP),soil environment(soil water content,electrical conductance,and pH),vegetation factors,and elevation to determine which of them are the main driving factors of the spatial variability of SMF in the Altun Shan.We explored the best-fit model of SMF along the changes in elevation using a structural equation model,performed variance partitioning analysis(VPA)on SMF with the“varpart”function to explain the relative contribution of various environmental factors to SMF changes,and used a random forest model for relative importance analysis.The results showed that SMF in the Altun Shan significantly increased with elevation in a linear trend.The main driver of changes in SMF was found to be MAP.Although the rise in elevation did not have a significant direct effect on changes in SMF,it could indirectly affect SMF by significantly influencing MAP,p H,MAT,and normalized difference vegetation index(NDVI).When considering climate,soil environment,and vegetation factors together,they explained 76%of the variation in SMF.The largest contribution to the variation in SMF was attributed to the independent effect of climate(0.31)and its interactive effect with soil(0.30).The relative importance of MAP on SMF changes was found to be the greatest.It is indicated that changes in SMF are caused by the combined effect of multiple environmental conditions.These findings are essential for understanding the spatial variability and drivers of SMF in dryland mountain ecosystems,especially concerning the function of mountain ecosystems in the context of global climatic changes.
文摘随着互联网上Mashup服务数量及种类的急剧增长,如何从这些海量的服务集合中快速、精准地发现满足用户需求的Mashup服务,成为一个具有挑战性的问题.针对这一问题,本文提出一种融合功能语义关联计算与密度峰值检测的Mashup服务聚类方法,用于缩小服务的搜索空间,提升服务发现的精度与效率.首先,该方法对Mashup服务进行元信息提取和描述文本内容整理,并根据Web API组合的标签对相应Mashup服务标签进行扩充.然后,用基于功能语义关联计算方法(Functional Semantic Association Calculation Method,FSAC)提取出各服务描述的功能名词集合,并通过功能名词的语义权重来构造Mashup语义特征向量.最后,通过基于密度信息的聚类中心检测方法(Clustering Center Detection Method based on Density Information,CCD-DI)检测出最为合适的K个Mashup语义特征向量作为K-means算法的初始中心,进行聚类划分.基于ProgrammableWeb的真实数据实验表明,本文所提聚类方法在纯度、精准率、召回率、熵等指标上均有良好表现.