摘要：This paper presents a winning solution for the CCKS-2020 financial event extraction task, where the goal is to identify event types, triggers and arguments in sentences across multiple event types. In this task, we focus on resolving two challenging problems (i.e., low resources and element overlapping) by proposing a joint learning framework, named SaltyFishes. We first formulate the event extraction task as a joint probability model. By sharing parameters in the model across different types, we can learn to adapt to low-resource events based on high-resource events. We further address the element overlapping problems by a mechanism of Conditional Layer Normalization, achieving even better extraction accuracy. The overall approach achieves an F1-score of 87.8% which ranks the first place in the competition.
摘要：In this work, we introduce an extra singlet pseudoscalar into the Type-III two Higgs doublet model (2HDM) which is supposed to solve a series of problems in the modern particle-cosmology. With existence of a light pseudoscalar, the h → µτ excess measured at CMS and as well as the (g − 2)µ anomaly could be simultaneously explained within certain parameter spaces that can also tolerate the data on the flavor-violating processes τ → µγ and Higgs decay gained at LHC. Within the same parameter spaces, the DM relic abundance is well accounted. Moreover, the recently observed Galactic Center gamma ray excess(GCE) is proposed to realize through dark matter(DM) pair annihilations, and in this work, the scenario of the annihilation being mediated by the pseudoscalar is also addressed.
摘要：Investigating the CKM matrix in different parametrization schemes, it is noticed that those schemes can be divided into a few groups where the sine values of the CP phase for each group are approximately equal. Using those relations, several approximate equalities among the elements of CKM matrix are established. Assuming them to be exact, there are infinite numbers of solutions and by choosing special values for the free parameters in those solutions, several textures presented in literature are obtained. The case can also be generalized to the PMNS matrix for the lepton sector. In parallel, several mixing textures are also derived by using presumed symmetries, amazingly, some of their forms are the same as what we obtained, but not all. It hints existence of a hidden symmetry which is broken in the practical world. The nature makes its own selection on the underlying symmetry and the way to break it, while we just guess what it is.
摘要：If they find some words closely related to their target task of titles or abstracts, which are definedIf they find some words closely related to their target task of titles or abstracts, which are definedas intelligent sensitive words in this paper, they will decide to read these resources in depth tofind more useful information. These sensitive words may refer to special persons, organisations,programmes, terms and so on. All of these sensitive words and their featured information areincluded in our object. Basing on the above processes, we have determined that some featurescould be used for automatic computation.Downloaded by [National Science Library] at 03:28 19 January 2015 Profiling science and innovation policy by object-based computing 587? Authority of source. If one piece of news is released in an official website, then this resource ismore reliable than other resources from no