The paper "Neural Relation Extraction with Selective Attention over Instances" addresses the challenge of distant supervised relation extraction, which often suffers from noisy and incorrect labeling data. The authors propose a sentence-level attention-based convolutional neural network (CNN) model to improve the accuracy of relation extraction. The model uses CNNs to embed the semantics of sentences and then applies sentence-level attention to dynamically reduce the weights of noisy instances. Experimental results on real-world datasets show that the proposed model significantly outperforms existing methods, achieving consistent improvements in relation extraction tasks. The key contributions of the paper include the ability to fully utilize informative sentences and the effective handling of wrong labeling through selective attention. The source code for the model is available at https://github.com/thunlp/NRE.The paper "Neural Relation Extraction with Selective Attention over Instances" addresses the challenge of distant supervised relation extraction, which often suffers from noisy and incorrect labeling data. The authors propose a sentence-level attention-based convolutional neural network (CNN) model to improve the accuracy of relation extraction. The model uses CNNs to embed the semantics of sentences and then applies sentence-level attention to dynamically reduce the weights of noisy instances. Experimental results on real-world datasets show that the proposed model significantly outperforms existing methods, achieving consistent improvements in relation extraction tasks. The key contributions of the paper include the ability to fully utilize informative sentences and the effective handling of wrong labeling through selective attention. The source code for the model is available at https://github.com/thunlp/NRE.