Recent Submissions

  • Self-Supervised Hypergraph Convolutional Networks for Session-based Recommendation

    Xia, Xin; Yin, Hongzhi; Yu, Junliang; Wang, Qinyong; Cui, Lizhen; Zhang, Xiangliang (arXiv, 2020-12-12) [Preprint]
    Session-based recommendation (SBR) focuses on next-item prediction at a certain time point. As user profiles are generally not available in this scenario, capturing the user intent lying in the item transitions plays a pivotal role. Recent graph neural networks (GNNs) based SBR methods regard the item transitions as pairwise relations, which neglect the complex high-order information among items. Hypergraph provides a natural way to capture beyond-pairwise relations, while its potential for SBR has remained unexplored. In this paper, we fill this gap by modeling session-based data as a hypergraph and then propose a dual channel hypergraph convolutional network -- DHCN to improve SBR. Moreover, to enhance hypergraph modeling, we innovatively integrate self-supervised learning into the training of our network by maximizing mutual information between the session representations learned via the two channels in DHCN, serving as an auxiliary task to improve the recommendation task. Extensive experiments on three benchmark datasets demonstrate the superiority of our model over the SOTA methods, and the ablation study validates the effectiveness and rationale of hypergraph modeling and self-supervised task. The implementation of our model is available via https://github.com/xiaxin1998/DHCN.
  • Cross-equalization of time-lapse seismic data using recurrent neural networks

    Alali, Abdullah A.; Kazei, Vladimir; Sun, Bingbing; Smith, Robert; Nivlet, Phlippe; Bakulin, Andrey; Alkalifah, Tariq (Society of Exploration Geophysicists, 2020-09-30) [Conference paper]
    Time-lapse seismic uses repetitive seismic surveys to monitor the fluid in the subsurface. Ideally, the time-lapse data should be identical except for at the target region (i.e., the reservoir), where the fluid changes occur. Unfortunately, it is almost impossible to have identical data for various reasons, such as the static changes in the near-surface or the varying positioning of sources and receivers between surveys. To increase the accuracy of the 4D signal and reduce the noise, we propose to process the time-lapse data using a machine-learning methodology. Specifically, we train a recurrent neural network (RNN) model to map the data from monitor to baseline. The learned RNN model would reveal 4D overburden changes. Therefore, the difference between the predicted baseline and the actual baseline data stets will represent the target signal. We validate the method on synthetic data and show the improvements of the 4D signal by imaging the reservoir and computing the normalized root mean square.
  • Unlocking features of locked-unlocked anionic polymerization

    Li, Cun; Leng, Xuefei; Han, Li; Bai, Hongyuan; Yang, Lincan; Li, Chao; Zhang, Songbo; Liu, Pibo; Ma, Hongwei (Polymer Chemistry, Royal Society of Chemistry (RSC), 2020) [Article]
    Due to unique characteristics of 1-(tri-isopropoxymethylsilylphenyl)-1phenylethylene (DPE-Si(OiPr)3), quantitative locked (the living species became dormant for endcapping with DPE-Si(OiPr)3) and unlocked (the dormant species regained the activity after adding the alkali metal alkoxides) anionic polymerization has been realized. In this work, the features of locked-unlocked anionic polymerization were carefully investigated with sequential feeding strategies. By combining the results from SEC, 1H-NMR, MALDI-TOF-MS and DFT calculations, the main features of this mechanism were revealed as follows: (1) it is a kinetically controlled process (kDS and kSS) due to the inherent features of different living species. Comparable initiation and propagation rates are required to ensure the simultaneous chain growth from unlocked living species; (2) the transformation between locked and unlocked species depends on the alkalinity of the alkali metal alkoxides; the higher the alkalinity, the higher the unlocking efficiency. And promisingly, these findings may contribute to the modulation of molecular weight distribution and facilitate the preparation of position-defined functionalized polymers in the future.
  • CCDC 1838597: Experimental Crystal Structure Determination : {1,3-bis[2,6-bis(propan-2-yl)phenyl]-2,3-dihydro-1H-imidazol-2-ylidene}-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-gold tetrahydrofuran solvate

    Zinser, Caroline M.; Nahra, Fady; Falivene, Laura; Brill, Marcel; Cordes, David B.; Slawin, Alexandra M. Z.; Cavallo, Luigi; Cazin, Catherine S. J.; Nolan, Steven P. (Cambridge Crystallographic Data Centre, 2019-05-22) [Dataset]
  • Supplementary material from "Beyond the visual: using metabarcoding to characterize the hidden reef cryptobiome"

    Carvalho, Susana; Aylagas, Eva; Villalobos, Rodrigo; Kattan, Yasser; Berumen, Michael L.; Pearman, John K. (figshare, 2019) [Dataset]
    In an era of coral reef degradation, our knowledge of ecological patterns in reefs is biased towards large conspicuous organisms. The majority of biodiversity, however, inhabits small cryptic spaces within the framework of the reef. To assess this biodiverse community, which we term the ‘reef cryptobiome’, we deployed 87 autonomous reef monitoring structures (ARMS), on 22 reefs across 16 degrees latitude of the Red Sea. Combining ARMS with metabarcoding of the mitochondrial cytochrome oxidase I gene, we reveal a rich community, including the identification of 14 metazoan phyla within 10 416 operational taxonomic units (OTUs). While mobile and sessile subsets were similarly structured along the basin, the main environmental driver was different (particulate organic matter and sea surface temperature, respectively). Distribution patterns of OTUs showed that only 1.5% were present in all reefs, while over half were present in a single reef. On both local and regional scales, the majority of OTUs were rare. The high heterogeneity in community patterns of the reef cryptobiome has implications for reef conservation. Understanding the biodiversity patterns of this critical component of reef functioning will enable a sound knowledge of how coral reefs will respond to future anthropogenic impacts.