基于知识图谱的推荐方法是推荐系统的研究热点之一,利用用户历史行为及物品特征在知识图谱结构化表示数据的辅助下解决推荐系统数据稀疏性及冷启动问题。但是用户的兴趣易受特定物品所影响,而知识图谱以结构化形式存储数据,实体与实体...基于知识图谱的推荐方法是推荐系统的研究热点之一,利用用户历史行为及物品特征在知识图谱结构化表示数据的辅助下解决推荐系统数据稀疏性及冷启动问题。但是用户的兴趣易受特定物品所影响,而知识图谱以结构化形式存储数据,实体与实体之间仅存在链路关系,这就导致了单纯利用知识图谱的推荐方法在点击率预测方面性能较差。基于此,提出一种基于局部影响力与深层偏好传播的推荐方法(local influence and deep preference propagation,LIDP),充分利用知识图谱结构化数据在偏好传播中存在实体影响力的优点。LIDP模型首先对知识图谱逐层偏好传播获取数据影响力权重并根据所获数据权重进行局部影响力计算;其次对局部影响力进行用户历史行为的兴趣增强表示进而获取用户表示;最后对用户表示与物品的向量表示进行内积操作以获取最终交互概率。LIDP模型在MovieLens-1M数据集上相比最优基准模型GNRF,AUC、ACC、MAE和F 1值分别提高了0.16%、0.52%、0.87%、0.21%;在Book-Crossing数据集上,这些提升分别为0.45%、2.14%、1.29%、0.93%。实验结果表明,LIDP模型能有效获取深层次用户兴趣偏好,在推荐系统中具有良好的性能和效果,可以为用户提供更好的个性化推荐服务。展开更多
In the context of climate change,research on extreme climates and disaster risk management has become a crucial component of climate change adaptation.Local communities,which have been facing extreme climates for a lo...In the context of climate change,research on extreme climates and disaster risk management has become a crucial component of climate change adaptation.Local communities,which have been facing extreme climates for a long time in their production and daily life,have developed some locally applicable traditional knowledge that has played an important role in their adaptation to extreme climate and disaster risk management.Therefore,this research aims to link Local knowledge(LK)to community extreme climate disaster risk management in order to construct a conceptual model.It then takes the extreme climate adaptation strategy of traditional nomads in a temperate grassland of China as an example to analyze the role of LK in extreme climate adaptation using the proposed theoretical framework.The main research objectives of this study are:(1)To construct a conceptual model to illustrate the relations among extreme climate events,risk management,LK,and farmers’adaptation strategies;(2)To apply the theoretical framework to a field case to reveal context-specific extreme climate adaptation mechanisms with LK as a critical component;(3)To test the framework and provide suggestions for the extreme climates adaptation,and the conservation of LK related to climate change adaptation.The results show that from the perspective of disaster risk management,local communities could manage extreme climates as a disaster risk through adaptation strategies formed from LK,because as a knowledge system,LK contains relevant knowledge covering the whole process of disaster risk management.展开更多
文摘基于知识图谱的推荐方法是推荐系统的研究热点之一,利用用户历史行为及物品特征在知识图谱结构化表示数据的辅助下解决推荐系统数据稀疏性及冷启动问题。但是用户的兴趣易受特定物品所影响,而知识图谱以结构化形式存储数据,实体与实体之间仅存在链路关系,这就导致了单纯利用知识图谱的推荐方法在点击率预测方面性能较差。基于此,提出一种基于局部影响力与深层偏好传播的推荐方法(local influence and deep preference propagation,LIDP),充分利用知识图谱结构化数据在偏好传播中存在实体影响力的优点。LIDP模型首先对知识图谱逐层偏好传播获取数据影响力权重并根据所获数据权重进行局部影响力计算;其次对局部影响力进行用户历史行为的兴趣增强表示进而获取用户表示;最后对用户表示与物品的向量表示进行内积操作以获取最终交互概率。LIDP模型在MovieLens-1M数据集上相比最优基准模型GNRF,AUC、ACC、MAE和F 1值分别提高了0.16%、0.52%、0.87%、0.21%;在Book-Crossing数据集上,这些提升分别为0.45%、2.14%、1.29%、0.93%。实验结果表明,LIDP模型能有效获取深层次用户兴趣偏好,在推荐系统中具有良好的性能和效果,可以为用户提供更好的个性化推荐服务。
基金The National Natural Science Foundation of China (42001194)。
文摘In the context of climate change,research on extreme climates and disaster risk management has become a crucial component of climate change adaptation.Local communities,which have been facing extreme climates for a long time in their production and daily life,have developed some locally applicable traditional knowledge that has played an important role in their adaptation to extreme climate and disaster risk management.Therefore,this research aims to link Local knowledge(LK)to community extreme climate disaster risk management in order to construct a conceptual model.It then takes the extreme climate adaptation strategy of traditional nomads in a temperate grassland of China as an example to analyze the role of LK in extreme climate adaptation using the proposed theoretical framework.The main research objectives of this study are:(1)To construct a conceptual model to illustrate the relations among extreme climate events,risk management,LK,and farmers’adaptation strategies;(2)To apply the theoretical framework to a field case to reveal context-specific extreme climate adaptation mechanisms with LK as a critical component;(3)To test the framework and provide suggestions for the extreme climates adaptation,and the conservation of LK related to climate change adaptation.The results show that from the perspective of disaster risk management,local communities could manage extreme climates as a disaster risk through adaptation strategies formed from LK,because as a knowledge system,LK contains relevant knowledge covering the whole process of disaster risk management.
