Guojun Cheng, Jiaqi Luo, Jiasheng Qian and Jibin Miao
Titanium nitride (TiN) nano-particles were subjected to graft modification by silane coupling agent (KH-570) via a direct blending method. The hydroxyl groups on the surface of TiN nano-particles can interact with silanol groups [-Si-OCH3] of KH-570 forming an organic coating layer. The covalent bonds (Ti-O-Si) formation was testified by Fourier transform infrared spectra (FTIR) and X-ray photoelectron spectroscopy (XPS). Through transmission electron micrograph (TEM) observations, it was found that KH-570 could improve the dispersibility of nano-TiN particles in ethyl acetate. Thermo gravimetric analysis (TGA) and contact angle measurements indicated that KH-570 molecules were adsorbed or anchored on the surface of nano-TiN particle and the net efficiency of it was 22.76 %, which facilitated to hinder the aggregation of nano-TiN particles.
Yukun Zheng, Yiqun Liu, Zhen Fan, Cheng Luo, Qingyao Ai, Min Zhang and Shaoping Ma
A number of deep neural networks have been proposed to improve the performance of document ranking in information retrieval studies. However, the training processes of these models usually need a large scale of labeled data, leading to data shortage becoming a major hindrance to the improvement of neural ranking models’ performances. Recently, several weakly supervised methods have been proposed to address this challenge with the help of heuristics or users’ interaction in the Search Engine Result Pages (SERPs) to generate weak relevance labels. In this work, we adopt two kinds of weakly supervised relevance, BM25-based relevance and click model-based relevance, and make a deep investigation into their differences in the training of neural ranking models. Experimental results show that BM25-based relevance helps models capture more exact matching signals, while click model-based relevance enhances the rankings of documents that may be preferred by users. We further proposed a cascade ranking framework to combine the two weakly supervised relevance, which significantly promotes the ranking performance of neural ranking models and outperforms the best result in the last NTCIR-13 We Want Web (WWW) task. This work reveals the potential of constructing better document retrieval systems based on multiple kinds of weak relevance signals.
Objective: To explore the effects and underlying mechanisms of Ge Hua Jie Cheng Decoction on rats with alcoholic liver injuries.
Methods: 60 Wistar rats were randomly assigned to six groups: normal group, model group, Yi Gan Ling group, and Ge Hua Jie Cheng Decoction groups in low, middle and high concentrations, 10 rats in each group. Except for the normal group, rats in other groups were administered white wine for eight weeks to establish the liver injury model. During the modeling, the Yi Gan Ling/Ge Hua Jie Cheng Decoction were administered intragastrically to the rats. So the histopathological changes were observed after eight weeks, meanwhile the serum γ- glutamyl endopeptidase (GGT), Glutathione (GSH) and aspartate aminotransferase mitochondrial isoenzyme (m-AST) were assayed by automatic biochemical analyzer.
Results: under the light microscope, the groups of high and middle dosages of Ge Hua Jie Cheng Decoction , especially the high one, had apparent improvement of inflammatory infiltration in liver tissues. Compared with the normal group, the serum GGT and m-AST levels had elevated (P＜0.01), whereas the serum GSH level decreased (P＜0.01); compared with model group, the high and middle dosages of Ge Hua Jie Cheng Decoction groups had decreased serum GGT and m-AST (P＜0.01 or P＜0.05), as well as increased serum GSH level (P＜0.01).
Conclusion: Ge Hua Jie Cheng Decoction has a protective effect for liver injuries induced by alcohol, and this effect is dose-dependent. The high dosage showed stronger protection effect, which might be related to the increased serum GSH and decreased serum GGT and m-AST.