In this paper, we develop a convolutional neural network for stance detection in tweets. According to the official results, our system ranks 1 st on subtask B (among 9 teams) and ranks 2 nd on subtask A (among 19 teams) on the twitter test set of SemEval2016 Task 6. The main contribution of our work is as follows. We design a "vote scheme" for prediction instead of predicting when the accuracy of validation set reaches its maximum. Besides, we make some improvement on the specific subtasks. For subtask A, we separate datasets into five sub-datasets according to their targets, and train and test five separate models. For subtask B, we establish a two-class training dataset from the official domain corpus, and then modify the softmax layer to perform three-class classification. Our system can be easily re-implemented and optimized for other related tasks.

Dynamic graphs such as the user-item interactions graphs and financial transaction networks are ubiquitous nowadays. While numerous representation learning methods for static graphs have been proposed, the study of dynamic graphs is still in its infancy. A main challenge of modeling dynamic graphs is how to effectively encode temporal and structural information into nonlinear and compact dynamic embeddings. To achieve this, we propose a principled graph-neural-based approach to learn continuous-time dynamic embeddings. We first define a temporal dependency interaction graph() that is induced from sequences of interaction data. Based on the topology of this , we develop a dynamic message passing neural network named TDIG-MPNN, which can capture the fine-grained global and local information on. In addition, to enhance the quality of continuous-time dynamic embeddings, a novel selection mechanism comprised of two successive steps, i.e., co-attention and gating, is applied before the above TDIG-MPNN layer to adjust the importance of the nodes by considering high-order correlation between interactive nodes'-depth neighbors on. Finally, we cast our learning problem in the framework of temporal point processes (TPPs) where we use TDIG-MPNN to design a neural intensity function for the dynamic interaction processes. Our model achieves superior performance over alternatives on temporal interaction prediction (including tranductive and inductive tasks) on multiple datasets.

Forty nitramines by incorporating −CO, −NH2, −N3, −NF2, −NHNO2, −NHNH2, −NO2, −ONO2, −C­(NO2)3, and −CH­(NO2)2 groups based on a 1,3,5,7-tetranitro-1,3,5,7-tetraazacyclooctane (HMX) framework were designed. Their electronic structures, heats of formation (HOFs), detonation properties, thermal stabilities, electrostatic potential, and thermodynamic properties were systematically investigated by density functional theory. The comprehensive relationships between the structures and performance of different substituents were studied. Results indicate that −C­(NO2)3 has the greatest effect on improvement of HOFs among the whole substituted groups. Thermodynamic parameters, such as standard molar heat capacity (C p,m θ), standard molar entropy (S m θ), and standard molar enthalpy (H m θ), of all designed compounds increase with the increasing number of energetic groups, and the volumes of energetic groups have a great influence on standard molar enthalpy. Except for −NH2(R1), −NHNH2(R5), and B3, all of the designed compounds have higher detonation velocities and pressures than HMX. Among them, E7 exhibits an extraordinarily high detonation performance (D = 10.89 km s–1, P = 57.3 GPa), E1 exhibits a relatively poor detonation performance (D = 8.93 km s–1, P = 35.5 GPa), and −NF2 and −C­(NO2)3 are the best ones in increasing the density by more or less 12%.

Tumor dormancy continues to be a research hotspot with numerous pressing problems that need to be solved. The goal of this study is to perform a bibliometric analysis of pertinent articles published in the twenty-first century. We concentrate on significant keywords, nations, authors, affiliations, journals, and literature in the field of tumor dormancy, which will help researchers to review the results that have been achieved and better understand the directions of future research. We retrieved research articles on tumor dormancy from the Web of Science Core Collection. This study made use of the visualization tools VOSviewer, CiteSpace, and Scimago Graphica, as visualization helps us to uncover the intrinsic connections between information. Research on tumor dormancy has been growing in the 21st century, especially from 2015 to the present. The United States is a leader in many aspects of this research area, such as in the number of publications, the number of partners, the most productive institutions, and the authors working in this field. Harvard University is the institution with the highest number of publications, and Aguirre-Ghiso, Julio A. is the author with the highest number of publications and citations. The keywords that emerged after 2017 were “early dissemination”, “inhibition”, “mechanism”, “bone metastasis”, and “promotion”. We believe that research on tumor dormancy mechanisms and therapy has been, and will continue to be, a major area of interest. The exploration of the tumor dormancy microenvironment and immunotherapeutic treatments for tumor dormancy is likely to represent the most popular future research topics.