分类: 计算机科学 >> 自然语言理解与机器翻译 分类: 图书馆学、情报学 >> 情报过程自动化的方法和设备 提交时间: 2019-10-29
摘要: Purpose: Move recognition in scientific abstracts is an NLP task of classifying sentences of the abstracts into different types of language unit. To improve the performance of move recognition in scientific abstracts, a novel model of move recognition is proposed that outperforms BERT-Base method. Design: Prevalent models based on BERT for sentence classification often classify sentences without considering the context of the sentences. In this paper, inspired by the BERT's Masked Language Model (MLM), we propose a novel model called Masked Sentence Model that integrates the content and contextual information of the sentences in move recognition. Experiments are conducted on the benchmark dataset PubMed 20K RCT in three steps. And then compare our model with HSLN-RNN, BERT-Base and SciBERT using the same dataset. Findings: Compared with BERT-Base and SciBERT model, the F1 score of our model outperforms them by 4.96% and 4.34% respectively, which shows the feasibility and effectiveness of the novel model and the result of our model comes closest to the state-of-the-art results of HSLN-RNN at present. Research Limitations: The sequential features of move labels are not considered, which might be one of the reasons why HSLN-RNN has better performance. And our model is restricted to dealing with bio-medical English literature because we use dataset from PubMed which is a typical bio-medical database to fine-tune our model. Practical implications: The proposed model is better and simpler in identifying move structure in scientific abstracts, and is worthy for text classification experiments to capture contextual features of sentences. Originality: The study proposes a Masked Sentence Model based on BERT which takes account of the contextual features of the sentences in abstracts in a new way. And the performance of this classification model is significantly improved by rebuilding the input layer without changing the structure of neural networks.
分类: 物理学 >> 普通物理:统计和量子力学,量子信息等 提交时间: 2017-05-02
摘要: Ordered tungsten and aluminum co-doped mesoporous KIT-6 catalysts (W-Al-KIT-6) with different Si/Al molar ratios were successfully synthesized by a one-pot synthesis method. The obtained W-Al-KIT-6 catalysts were tested for catalytic conversion of 1-butene and ethene to propene via isomerization of 1-butene to 2-butene and subsequent cross metathesis of 2-butene and ethene. Various characterization techniques, such as ICP-OES, XRD, BET, TEM, Raman, XPS and NH3-TPD, were used to characterize the catalysts. The introduction of Al did not change the mesoporous structure of KIT-6 when the nominal Si/Al was 10, 20 or 30. Moreover, the sample demonstrated a larger amount of acidic sites. The W-Al-KIT-6 catalysts with suitable Si/Al ratios illustrated a superior catalytic performance to W-KIT-6 catalyst. The origin of catalytic performance enhancement over W-Al-KIT-6 catalysts is preliminarily discussed and ascribed to the highly disperse W species and a large amount of acidic sites. The acidic sites were formed by the introduction of a suitable amount of Al in the W-KIT-6 framework, which accelerated the isomerization of 1-butene to 2-butene and promoted the cross metathesis of 2-butene and ethene to propene.
分类: 物理学 >> 普通物理:统计和量子力学,量子信息等 提交时间: 2017-05-08
摘要: Tungsten substituted mesoporous FDU-12 (W-FDU-12) catalysts were synthesized by a one-pot hydrothermal process using F127 as the structure directing agent. The studies of TEM, SAXS and BET illustrated that the highly ordered mesoporous structure of FDU-12 was maintained in the doped W-FDU-12 samples. XPS studies revealed that a high concentration of W5+ species appeared in doped W-FDU-12 catalysts whereas supported WO3/FDU-12 and WO3/SiO2 catalysts only contained W6+ species. Tandem catalytic conversion of 1-butene and ethene to propene through isomerization of 1-butene to 2-butene and consecutive cross metathesis of 2-butene and ethene in a fixed-bed reactor at different temperatures and atmospheric pressure was used to evaluate the catalytic performance of the W-FDU-12 catalyst, combined with MgO. The catalytic results showed that the doped W-FDU-12 illustrated a superior catalytic performance relative to the supported WO3/FDU-12 and WO3/SiO2 catalysts. The higher metathesis activity of W-FDU-12 catalysts can be ascribed to the good dispersion of W species and the incorporation of W species into the framework of FDU-12, forming a substantial amount of W5+, which was beneficial for the cross metathesis of 2-butene and ethene to propene.
分类: 生物学 >> 生物物理学 提交时间: 2016-05-12
摘要: The phosphoinositide 3-kinase (PI3K)/protein kinase B (AKT) pathway has been identified as an important pathway in renal cell carcinoma (RCC). We have reported a nonsense mutation in PIK3R1, which encodes the regulatory subunit of PI3K, in a metastatic RCC (mRCC), while the mutation was absent in the corresponding primary RCC (pRCC). To identify the function of PIK3R1 in RCC, we examined its expression in normal kidney, pRCC and mRCC by immunohistochemistry and real-time polymerase chain reaction. The expression of PIK3R1 significantly decreased in pRCC and was further reduced in mRCC compared with normal tissue. Besides, its expression levels were negatively correlated with T-category of tumor stage. Additionally, 786-O and A-704 cells with PIK3R1 depletion introduced by CRISPR/Cas9 system displayed enhanced proliferation, migration and epithelial-mesenchymal transition (EMT), and acquired a stem-like phenotype. Moreover, the PIK3R1 depletion promoted the phosphorylation of AKT in the cells. The knockdown of AKT by shRNA reduced p-GSK3 beta and CTNNB1 expression in the cells, while the depletion of CTNNB1 impaired stem-like phenotype of the cells. Overall, PIK3R1 down-regulation in RCC promotes propagation, migration, EMT and stem-like phenotype in renal cancer cells through the AKT/GSK3 beta/CTNNB1 pathway, and may contribute to progression and metastasis of RCC.