Several outbreaks of COVID-19 were associated with seafood markets, raising concerns that fish-attached SARS-CoV-2 may exhibit prolonged survival in low-temperature environments. Here we showed that salmon-attached SARS-CoV-2 at 4oC could remain infectious for more than one week, suggesting that fish-attached SARS-CoV-2 may be a source of transmission.
Many deep learning architectures have been employed to model the semantic compositionality for text sequences, requiring a huge amount of supervised data for parameters training, making it unfeasible in situations where numerous annotated samples are not available or even do not exist. Different from data-hungry deep models, lightweight word embedding-based models could represent text sequences in a plug-and-play way due to their parameter-free property. In this paper, a modified hierarchical pooling strategy over pre-trained word embeddings is proposed for text classification in a few-shot transfer learning way. The model leverages and transfers knowledge obtained from some source domains to recognize and classify the unseen text sequences with just a handful of support examples in the target problem domain. The extensive experiments on five datasets including both English and Chinese text demonstrate that the simple word embedding-based models (SWEMs) with parameter-free pooling operations are able to abstract and represent the semantic text. The proposed modified hierarchical pooling method exhibits significant classification performance in the few-shot transfer learning tasks compared with other alternative methods. INDEX TERMS Few-shot learning, transfer learning, text classification, word embedding based models, pooling strategy.
Service-oriented architecture (SOA) provides the concept of packaging available functionalities as interoperable services within the context of various domains that use it. With rapid advances in SOA technologies and the growing availability of web services, the problem of composing a set of web services to achieve complex systems is becoming more practical. In the context of cyber-physical systems where hardware and software are coupled together to realize integrated systems, there are special characteristics and requirements. One of these is that most service providers are physical entities with their own states and properties. The constraint that follows is that a given entity might not be able to perform all the services it can provide at the same time. In fact, "multi-threading" for physical entities is rarely possible. This requires specific service modeling techniques to enable the use of SOA methods for this domain. Another characteristic is that due to the dynamic nature of cyber-physical worlds, the service composition procedure must be dynamically adaptive. In terms of AI planning, which is one of the fundamental techniques for service composition, not only the initial state and the goal are dynamic, but the planning domain also needs to be generated dynamically to provide the complete input to the underlying planner.Taking all these characteristics and requirements into consideration, we develop an ontology model for physical entity specification. Based on this model, another widely used service ontology model, OWL-S is extended to accurately model the characteristics of service providers in the context of cyber-physical worlds. Further, a technique for generating planning domain based on task requirements is developed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.