No abstract

Considering the natural tendency of people to follow direct or indirect cues of other people's activities, collaborative filtering-based recommender systems often predict the utility of an item for a particular user according to previous ratings by other similar users. Consequently, effective searching for the most related neighbors is critical for the success of the recommendations. In recent years, collaborative tagging systems with social bookmarking as their key component from the suite of Web 2.0 technologies allow users to freely bookmark and assign semantic descriptions to various shared resources on the web. While the list of favorite web pages indicates the interests or taste of each user, the assigned tags can further provide useful hints about what a user thinks of the pages.In this paper, we propose a new collaborative filtering approach TBCF (Tag-based Collaborative Filtering) based on the semantic distance among tags assigned by different users to improve the effectiveness of neighbor selection. That is, two users could be considered similar not only if they rated the items similarly, but also if they have similar cognitions over these items. We tested TBCF on real-life datasets, and the experimental results show that our approach has significant improvement against the traditional cosine-based recommendation method while leveraging user input not explicitly targeting the recommendation system.

Considering the heavy burden of migraine, it is essential to update insufficient and/or outdated clinical evidence supporting electroacupuncture (EA) in migraine therapy. In this study, a literature search of seven medical databases was performed. After data extraction and quality evaluation, 13 randomized controlled trials, including 1559 patients, were assessed in this analysis. Results demonstrated that EA was superior to control treatment (Western medicine, sham-EA, blank control, acupuncture, and acupoint catgut embedding) according to the visual analog scale (VAS) score, frequency of headache attack (Western medicine, sham-EA, blank control), self-rating anxiety scale (SAS [blank control]), self-rating depression score (SDS [Western medicine and blank control]), and clinical efficiency (Western medicine and sham-EA) after treatment ([Formula: see text]). Results of network meta-analysis (for VAS, SAS, and SDS) demonstrated statistically significant differences in VAS scores for EA compared with sham-EA, acupuncture with sham-EA, acupoint catgut embedding with sham-EA, and acupoint catgut embedding with blank control. Rank probability analysis of VAS, SAS, and SDS scores all demonstrated that EA ranked first. Most studies were symmetrically distributed on both sides of the midline in funnel plots for VAS, SAS, and SDS, which indicated a low likelihood of small sample effects. Sensitivity analysis confirmed the stability of the studies included in this research. EA is one of several effective treatments for migraine pain symptoms, and, to some extent, anxiety and depression. Nevertheless, multi-center studies with large sample sizes and/or well-designed randomized controlled trials (RCTs) will be needed in the future.

Object. The purpose of this study was to fully assess the role of statins in the primary prevention of coronary heart disease (CHD). Methods. We searched six databases (PubMed, the Cochrane Library, Web of Science, China National Knowledge Infrastructure, Wanfang Database, and Chinese Scientific Journal Database) to identify relevant randomized controlled trials (RCTs) from inception to 31 October 2017. Two review authors independently assessed the methodological quality and analysed the data using Rev Man 5.3 software. Risk ratios and 95% confidence intervals (95% CI) were pooled using fixed/random-effects models. Funnel plots and Begg’s test were conducted to assess publication bias. The quality of the evidence was evaluated using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach. Results. Sixteen RCTs with 69159 participants were included in this review. Statins can effectively decrease the occurrence of angina (RR=0.70, 95% CI: 0.58~0.85, I2 =0%), nonfatal myocardial infarction (MI) (RR=0.60, 95% CI: 0.51~0.69, I2 =14%), fatal MI (RR=0.49, 95% CI: 0.24~0.98, I2 =0%), any MI (RR=0.53, 95% CI: 0.42~0.67, I2 =0%), any coronary heart events (RR=0.73, 95% CI: 0.68~0.78, I2=0%), coronary revascularization (RR=0.66, 95% CI: 0.55~0.78, I2 = 0%), and any cardiovascular events (RR=0.77, 95% CI: 0.72~82, I2 = 0%). However, based on the current evidence, there were no significant differences in CHD deaths (RR=0.82, 95% CI: 0.66~1.02, I2=0%) and all-cause mortality (RR=0.88, 95% CI: 0.76 ~1.01, I2 =58%) between the two groups. Additionally, statins were more likely to result in diabetes (RR=1.21, 95% CI: 1.05~1.39, I2 =0%). There was no evidence of publication biases, and the quality of the evidence was considered moderate. Conclusion. Statins seemed to be beneficial for the primary prevention of CHDs but have no effect on CHD death and all-cause mortality.

Multi-label classification, or the same example can belong to more than one class label, happens in many applications. To name a few, image and video annotation, functional genomics, social network annotation and text categorization are some typical applications. Existing methods have limited performance in both efficiency and accuracy. In this paper, we propose an extension over decision tree ensembles that can handle both challenges. We formally analyze the learning risk of Random Decision Tree (RDT) and derive that the upper bound of risk is stable and lower bound decreases as the number of trees increases. Importantly, we demonstrate that the training complexity is independent from the number of class labels, a significant overhead for many state-of-the-art multi-label methods. This is particularly important for problems with large number of multi-class labels. Based on these characteristics, we adopt and improve RDT for multi-label classification. Experiment results have demonstrated that the computation time of the proposed approaches is 1-3 orders of magnitude less than other methods when handling datasets with large number of instances and labels, as well as improvement up to more than 10% in accuracy as compared to a number of state-of-the-art methods in some datasets for multi-label learning. Considering efficiency and effectiveness together, Multi-label RDT is the top rank algorithm in this domain. Even compared with the HOMER algorithm proposed to solve the problem of large number of labels, Multi-label RDT runs 2-3 orders of magnitude faster in training process and achieves some improvement on accuracy. Software and datasets are available from the authors.