Falcon: A Novel Chinese Short Text Classification Method

For natural language processing problems, the short text classification is still a research hot topic, with obviously problem in the features sparse, high-dimensional text data and feature representation. In order to express text directly, a simple but new variation which employs one-hot with low-dimension was proposed. In this paper, a Densenet-based model was proposed to short text classification. Furthermore, the feature diversity and reuse were implemented by the concat and average shuffle operation between Resnet and Densenet for enlarging short text feature selection. Finally, some benchmarks were introduced to evaluate the Falcon. From our experimental results, the Falcon method obtained significant improvements in the state-of-art models on most of them in all respects, especially in the first experiment of error rate. To sum up, the Falcon is an efficient and economical model, whilst requiring less computation to achieve high performance.

Enzymatic pretreatment mitigates the dissemination of antibiotic resistance genes via regulating microbial populations and gene expressions during food waste fermentation

Food waste(FW)has been recognized as essential reservoir for resource recovery via anaerobic fermentation,which could also bring the potential risk of antibiotic resistance genes(ARGs)dissemination.Although the structural deficiency of FW could be stimulated by enzymatic pretreatment to enhance fermentation efficiency,the influences of enzymatic pretreatment on ARGs fate and microbial metabolic pathways involved in ARGs dissemination have rarely been reported.This work proved that enzymatic pretreatment could effectively decrease the total abundance of ARGs(reduced by 13.8%-24.5%)during long-term FW fermentation.It was found that enzymatic pretreatment significantly reduced the ARGs belonging to the efflux pump,which might be ascribed to its ability to increase membrane permeability.Furthermore,enzymatic pretreatment was in favor of reducing microbial diversity and various potential ARGs host(e.g.,Methanosarcina,Clostridium,Prevotella,Parabacteroides).Also,this pretreatment remarkably up-regulated the genetic expressions involved in ABC transporter(e.g.,eryF and mntA)and down-regulated the genetic expressions that participated in DNA replication,two-component systems(e.g.,uphA and cckA),and quorum sensing(e.g.,rpfF and lsrG),thereby decreasing ARGs transmission.This study would expand the insight of the influences of pretreatment method on ARGs fate during FW fermentation,and offer practical guidance on the sustainable management of FW.

A Plant Immune Receptor Adopts a Two-Step Recognition Mechanism to Enhance Viral Effector Perception

Plant intracellular nucleotide binding leucine-rich repeat (NLR) immune receptors play critical roles in pathoge n surveillance. Most plant NLRs characterized so far were found to use a single domain/sensor to recognize pathogen effectors. Here we report that the Sw-5b NLR immune receptor uses two distinct domains to detect the viral movement protein NSm encoded by tospovirus. In addition to its leucine-rich repeat (LRR) domain that has been previously reported, the N-terminal Solanaceae domain (SD) of Sw- 5b also interacts with NSm and a conserved 21-amino-acid region of NSm (NSm^21). The specific interaction between Sw-5b SD and NSm is required for releasing the inhibitory effect of coiled-coil domain on the NBARC- LRR region. Furthermore, we found that the binding of NSm affects the nucleotide binding activity of the NB-ARC-LRR in vitro, while Sw-5b NB-ARC-LRR is activated only when NSm and NSm^21 levels are high. Interestingly, Sw-5b SD could significantly enhanee the ability of the NB-ARC-LRR to detect low levels of NSm effector and facilitate its activation and induction of defense response. An Sw-5b SD mutant that is disrupted in NSm recognition failed to enhance the ability of the NB-ARC-LRR to sense low levels of NSm and NSm^21 . Taken together, our results suggest that Sw-5b SD functions as an extra sensor and the NB-ARC-LRR as an activator, and that Sw-5b NLR adopts a two-step recog nition mechanism to enhance viral effector perception.

A Sequence-Scheduled and Query-Based MAC Protocol for Underwater Acoustic Networks with a Mobile Node

Since mobile nodes such as autonomous underwater vehicles can effectively expand network coverage,underwater acoustic networks with mobile nodes have attracted more and more attention in recent years.To ensure the timely access of a mobile node without disturbing the normal transmission of data packets from static nodes,a sequence-scheduled and query-based medium access control protocol is proposed in this paper.The underwater nodes with data packets to be transmitted reply with a control packet for channel reservation after receiving the broadcast control packet of the buoy,then the reserved nodes transmit their data packets as the order calculated by the buoy.We develop a mechanism to adjust the transmission order of the control packets sent by the underwater nodes in the handshake phase to reduce the collision and guarantee the success rate of reservation,and a query is initiated when the reserved control packet of the mobile node is not received by the buoy in the handshake to respond to the data transmission request of the mobile node in time.Simulations show that the maximum nodal throughput of the proposed MAC protocol increases by at least 20%and the access delay of the mobile node decreases by about 90%,compared to the two reference protocols.Besides,the average access probability of the mobile node in the proposed protocol is more than 90%.