Search

• We take advantage of recent distributed word representation technology to help summarize text chunks. • The word embeddings in this research are utilized both in semantically and numerically. • We explore the relationships between natural language phrases in semantic view to extract latent knowledge domains. • Users' expertise is determined on historical performance, and data sparseness is alleviated by matrix factorization. • Evaluation is conducted on large scale dataset by comprehensive experiments, where preferable output is generated. Community Question Answering (CQA) websites can be claimed as the most major venues for knowledge sharing, and the most effective way of exchanging knowledge at present. Considering that massive amount of users are participating online and generating huge amount data, management of knowledge here systematically can be challenging. Expert recommendation is one of the major challenges, as it highlights users in CQA with potential expertise, which may help match unresolved questions with existing high quality answers while at the same time may help external services like human resource systems as another reference to evaluate their candidates. In this paper, we in this work we propose to exploring experts in CQA websites. We take advantage of recent distributed word representation technology to help summarize text chunks, and in a semantic view exploiting the relationships between natural language phrases to extract latent knowledge domains. By domains, the users' expertise is determined on their historical performance, and a rank can be compute to given recommendation accordingly. In particular, Stack Overflow is chosen as our dataset to test and evaluate our work, where inclusive experiment shows our competence. [ABSTRACT FROM AUTHOR]

The Coreopsis tinctoria Nutt. flower (CTF) has been used traditionally in China for treating hypertension and diabetes as well as reducing body weight and blood fat. However, the vascular protection effect of the CTF has not been studied to date. This study aimed to screen and identify bioactive fractions from the CTF with a diabetic endothelial protection effect and to clarify the underlying mechanism. The vascular protection effect of Fraction A was studied in high-fat diet and streptozocin-induced diabetic models. The endothelial protection effect of Fraction A-2 was further studied in an in vitro vascular endothelial dysfunction model induced by high glucose. In a high glucose-induced human umbilical vein endothelial cell (HUVEC) model, Fractions A-2-2 and A-2-3 were screened, and their detailed mechanisms of endothelial protection were studied. Liquid chromatography mass spectrometry (LC-MS) was used to identify the main components in Fractions A-2-2 and A-2-3. Fraction A treatment significantly improved the endothelium-dependent vasodilation of the mesenteric artery induced by acetylcholine in diabetic rats. The maximum relaxation was 79.82 ± 2.45% in the control group, 64.36 ± 9.81% in the model group, and 91.87 ± 7.38% in the Fraction A treatment group (P < 0.01). Fraction A treatment also decreased rat tail pressure compared with the model group at the 12th week. The systolic blood pressure was 152.7 5 ± 16.99 mmHg in the control group, 188.50 ± 5.94 mmHg in the model group, and 172.60 ± 14.31 mmHg in the Fraction A treatment group (P < 0.05). The mean blood pressure was 128.50 ± 13.79 mmHg in the control group, 157.00 ± 6.06 mmHg in the model group, and 144.80 ± 11.97 mmHg in the Fraction A treatment group (P < 0.05). In an in vitro vascular endothelium-dependent vasodilation dysfunction model induced by high glucose, Fraction A-2 improved the vasodilation of the mesenteric artery. The maximum relaxation was 82.15 ± 16.24% in the control group, 73.29 ± 14.25% in the model group, and 79.62 ± 13.89% in the Fraction A-2 treatment group (P < 0.05). In a high glucose-induced HUVEC model, Fraction A-2-2 and Fraction A-2-3 upregulated the expression of IRS-1, Akt, and eNOS and increased the levels of p-IRS-1Ser307, p-Akt Ser473, and p-eNOSSer1177 and also decreased the expression of NOX4, TNF-α, IL-6, sVCAM, sICAM, and NF-κB (P < 0.01). With the intervention of AG490 and LY294002, the above effects of Fraction A-2-2 and Fraction A-2-3 were inhibited (P < 0.01). LC-MS data showed that in Fraction A-2-2 and Fraction A-2-3, there were 10 main components: flavanocorepsin; polyphenolic; flavanomarein; isochlorogenic acid A; dicaffeoylquinic acid; coreopsin; marein; coreopsin; luteolin-7-O-glucoside; and 3′,5,5',7-tetrahydroxyflavanone-O-hexoside. The protective effect of the CTF on diabetic endothelial dysfunction may be due to its effect on the JAK2/IRS-1/PI3K/Akt/eNOS pathway and the related oxidative stress and inflammation. The results strongly suggested that Fraction A-2-2 and Fraction A-2-3 were the active fractions from the CTF, and the CTF might be a potential option for the prevention of vascular complications in diabetes. Image 1 [ABSTRACT FROM AUTHOR]

• We employ statistical analysis to define a list of quality features. • We propose a joint model by fusing autoencoder and random forest in an end-to-end trainable way for fake review detection. • Experimental results on the real dataset demonstrate the efficiency of the model. Online reviews play an important role in influencing buyers' daily purchase decisions. However, fake and meaningless reviews, which cannot reflect users' genuine purchase experience and opinions, widely exist on the Web and pose great challenges for users to make right choices. Therefore, it is desirable to build a fair model that evaluates the quality of products by distinguishing spamming reviews. We present an end-to-end trainable unified model to leverage the appealing properties from Autoencoder and random forest. A stochastic decision tree model is implemented to guide the global parameter learning process. Extensive experiments were conducted on a large Amazon review dataset. The proposed model consistently outperforms a series of compared methods. [ABSTRACT FROM AUTHOR]