Search

Sequence labeling is a crucial task in the NLP community that aims at identifying and assigning spans within the input sentence. It has wide applications in various fields such as information extraction, dialogue system, and sentiment analysis. However, previously proposed span-based or sequence-to-sequence models conduct locating and assigning in order, resulting in problems of error propagation and unnecessary training loss, respectively. This paper addresses the problem by reformulating the sequence labeling as a non-autoregressive set generation to realize locating and assigning in parallel. Herein, we propose a Dual-Query Set Generation (DQSetGen) model for unified sequence labeling tasks. Specifically, the dual-query set, including a prompted type query and a positional query with anchor span, is fed into the non-autoregressive decoder to probe the spans which correspond to the positional query and have similar patterns with the type query. By avoiding the autoregressive nature of previous approaches, our method significantly improves efficiency and reduces error propagation. Experimental results illustrate that our approach can obtain superior performance on 5 sub-tasks across 11 benchmark datasets. The non-autoregressive nature of our method allows for parallel computation, achieving faster inference speed than compared baselines. In conclusion, our proposed non-autoregressive dual-query set generation method offers a more efficient and accurate approach to sequence labeling tasks in NLP. Its advantages in terms of performance and efficiency make it a promising solution for various applications in data mining and other related fields.

Microglia, which are the primary inflammatory cells of the brain, can undergo phenotypic switching between M1 and M2 polarization, which have opposing effects on inflammation. Peroxisome proliferator-activated receptor gamma (PPARγ) is a member of the nuclear receptor family of ligand-inducible transcription factors, and PPARγ is known to regulate M2 macrophage polarization. Previous studies have shown that the natural pentacyclic triterpenoid ursolic acid (3β-hydroxy-urs-12-en-28-oic acid; UA) influences microglial activation. Additionally, UA increases tissue inhibitor matrix metalloproteinase 1 (TIMP1), while greatly reducing the release of matrix metalloproteinase 2 (MMP2) and MMP9 in a PPARγ-dependent manner. Here, we examined the anti-inflammatory properties of UA by observing how well it promotes the phenotypic transition of lipopolysaccharide (LPS) and interferon gamma (IFNγ)-activated BV2 microglia from M1 to M2 polarization. To determine if PPARγ is involved in the underlying molecular pathway, we treated rats with UA and the PPARγ inhibitor BADGE. We also investigated the mechanisms by which PPARγ controls transcription from the MMP2promoter. The in-vitro experiments showed that UA shifted LPS/IFNγ-activated BV2 microglia from the M1 to the M2 phenotype, which was associated with a reduction in the neurotoxic factors MMP2 and MMP9, and an increase in the anti-inflammatory factor TIMP1. Co-treatment with increased MMP2 and MMP9 synthesis while decreasing TIMP1 release, indicating that UA has anti-inflammatory effects on LPS/IFNγ-activated BV2 cells via activation of PPARγ. Next, we found that PPARγ directly influences MMP2 transcriptional activity by identifying the crucial peroxisome proliferator response element (PPRE) among five potential PPREs in the MMP2promoter. These results suggest that UA has a protective anti-inflammatory effect against neuroinflammatory toxicity, which is exerted by direct activation of PPARγ and selectively modulates microglial polarization and suppresses MMP2 formation.

Natural language generation from structured data mainly focuses on surface-level descriptions, suffering from uncontrollable content selection and low fidelity. Previous works leverage logical forms to facilitate logical knowledge-conditioned text generation. Though achieving remarkable progress, they are data-hungry, which makes the adoption for real-world applications challenging with limited data. To this end, this paper proposes a unified framework for logical knowledge-conditioned text generation in the few-shot setting. With only a few seeds logical forms (e.g., 20/100 shot), our approach leverages self-training and samples pseudo logical forms based on content and structure consistency. Experimental results demonstrate that our approach can obtain better few-shot performance than baselines.

Information extraction tasks such as entity relation extraction and event extraction are of great importance for natural language processing and knowledge graph construction. In this paper, we revisit the end-to-end information extraction task for sequence generation. Since generative information extraction may struggle to capture long-term dependencies and generate unfaithful triples, we introduce a novel model, contrastive information extraction with a generative transformer. Specifically, we introduce a single shared transformer module for an encoder-decoder-based generation. To generate faithful results, we propose a novel triplet contrastive training object. Moreover, we introduce two mechanisms to further improve model performance (i.e., batch-wise dynamic attention-masking and triple-wise calibration). Experimental results on five datasets (i.e., NYT, WebNLG, MIE, ACE-2005, and MUC-4) show that our approach achieves better performance than baselines.1

Objective Given its effects on lipid metabolism, the apolipoprotein A1-C3-A5 (APOA1-C3-A5) gene cluster is thought to play an important role in ischemic stroke pathogenesis. Here, we evaluated whether the APOA1-C3-A5cluster is associated with ischemic stroke in the northern Chinese Han population.Methods This case–control study analyzed 812 patients with ischemic stroke and 844 healthy controls with regard to four APOA1-C3-A5cluster promoter single nucleotide polymorphisms (SNPs), rs670, rs2854116, rs2854117, and rs662799, using the SNaPshot Multiplex sequencing assay. Potential associations among ischemic stroke, genotyping, and allele frequencies were assessed.Results APOA1rs670 CT/TT genotypes, APOA5rs662799 AG/GG genotypes, and the APOC3rs2854116 CC genotype were associated with an increased risk of ischemic stroke according to multivariate logistic analysis after adjusting for confounding factors. A significantly increased risk for ischemic stroke was also identified among high-risk haplotypes (C-C-T-A and T-T-C-A) for rs670–rs2854116–rs2854117–rs662799.Conclusion This study showed that rs670, rs2854116, and rs662799 SNPs of the APOA1-C3-A5cluster are associated with ischemic stroke in the northern Chinese Han population.

Objective The incidence of stroke has been rising annually and investigations into traditional risk factors have led to increased attention on genetic factors. In this study, we focused on the pri-let-7f gene, and investigated the association between pri-let-7f gene polymorphisms and ischemic stroke (IS).Methods This case–control study included 1803 patients and 1456 healthy controls of Han ethnicity living in Liaoning Province. We carried out genotyping analysis of two loci, pri-let-7f-1 rs10739971 and pri-let-7f-2 rs17276588, and performed statistical analysis controlling for confounding factors by logistic regression.Results The A alleles and AA genotypes of both loci were significantly associated with an increased risk of IS. Variant genotypes of rs17276588 may also increase the risk of IS in females with alcohol intake. Gene–gene interaction analysis showed combined effects of mutations in both these single nucleotide polymorphisms (SNPs).Conclusions This study demonstrated an association between pri-let-7f SNPs and IS, providing potential latent biomarkers for the risk of IS. However, more detailed studies are needed to clarify these results.

Knowledge graphs (KGs) express relationships between entity pairs, and many real-life problems can be formulated as knowledge graph reasoning (KGR). Conventional approaches to KGR have achieved promising performance but still have some drawbacks. On the one hand, most KGR methods focus only on one phase of the KG lifecycle, such as KG completion or refinement, while ignoring reasoning over other stages, such as KG extraction. On the other hand, traditional KGR methods, broadly categorized as symbolic and neural, are unable to balance both scalability and interpretability. To resolve these two problems, we take a more comprehensive perspective of KGR with regard to the whole KG lifecycle, including KG extraction, completion, and refinement, which correspond to three subtasks: knowledge extraction, relational reasoning, and inconsistency checking. In addition, we propose the implementation of KGR using a novel neural symbolic framework, with regard to both scalability and interpretability. Experimental results demonstrate that our proposed methods outperform traditional neural symbolic models.

The objective of this study was to evaluate the relationship of doctors’ job satisfaction with doctor-patient relationship and work-family conflict in China. The data came from a cross-sectional survey in Hubei province, which was part of China’s Fifth National Health Services Survey conducted in 2013. The survey in Hubei covered 54 secondary and tertiary general hospitals distributed in 20 counties. Of the 1080 questionnaires, 908 were included into our analysis. After surviving from reliability and validity tests, structural equation modeling was applied for further analysis with SPSS 20.0 and Mplus 7.0. The results showed that the average score of job satisfaction is 19.61 out of 30 points, indicating a relatively low level of doctors’ job satisfaction in Hubei province. Work-family conflict was found to have negative impact on doctors’ job satisfaction, and good doctor-patient relationship was found to have positive impact on their job satisfaction. Therefore, hospital administrators and policy makers should make effort to design and implement strategies that focus on meliorating doctor-patient relationship and balancing doctors’ work and family life to further improve their job satisfaction.