Online social network information promises to increase recommendation accuracy beyond the capabilities of purely rating/feedback-driven recommender systems (RS). As to better serve users' activities across different domains, many online social networks now support a new feature of "Friends Circles", which refines the domain-oblivious "Friends" concept. RS should also benefit from domain-specific "Trust Circles". Intuitively, a user may trust different subsets of friends regarding different domains. Unfortunately, in most existing multi-category rating datasets, a user's social connections from all categories are mixed together. This paper presents an effort to develop circle-based RS. We focus on inferring category-specific social trust circles from available rating data combined with social network data. We outline several variants of weighting friends within circles based on their inferred expertise levels. Through experiments on publicly available data, we demonstrate that the proposed circle-based recommendation models can better utilize user's social trust information, resulting in increased recommendation accuracy.

BackgroundAlthough evidence suggests that poor sleep is associated with chronic disease, little research has been conducted to assess the relationships between insufficient sleep, frequent mental distress (FMD ≥14 days during the past 30 days), obesity, and chronic disease including diabetes mellitus, coronary heart disease, stroke, high blood pressure, asthma, and arthritis.MethodsData from 375,653 US adults aged ≥ 18 years in the 2009 Behavioral Risk Factor Surveillance System were used to assess the relationships between insufficient sleep and chronic disease. The relationships were further examined using a multivariate logistic regression model after controlling for age, sex, race/ethnicity, education, and potential mediators (FMD and obesity).ResultsThe overall prevalence of insufficient sleep during the past 30 days was 10.4% for all 30 days, 17.0% for 14–29 days, 42.0% for 1–13 days, and 30.6% for zero day. The positive relationships between insufficient sleep and each of the six chronic disease were significant (p < 0.0001) after adjustment for covariates and were modestly attenuated but not fully explained by FMD. The relationships between insufficient sleep and both diabetes and high blood pressure were also modestly attenuated but not fully explained by obesity.ConclusionsAssessment of sleep quantity and quality and additional efforts to encourage optimal sleep and sleep health should be considered in routine medical examinations. Ongoing research designed to test treatments for obesity, mental distress, or various chronic diseases should also consider assessing the impact of these treatments on sleep health.

No abstract

Networks and the embedded relationships are critical determinants of how people communicate and form beliefs. The explosion of social media has significantly increased the scope and impact of social learning among consumers. This paper studies observational learning in networks of friends versus strangers. A consumer decides whether to adopt a product after receiving a private signal about product quality and observing the actions of others. The preference for the product has greater heterogeneity in the stranger-network than in the friend-network. We show that when the network is small, observing friends’ actions helps the consumer make more accurate inferences about quality. As the network grows, however, the stranger-network becomes more effective. Underlying these results are two competing effects of network heterogeneity on social learning. These are the individual preference effect, which allows one to make a better quality judgment when the preference element of past actions is more certain, and the social conforming effect, wherein private signals are underused in quality judgment as people follow others’ actions. We find cascading is more likely to occur in the friend-network than in the stranger-network. For a high-quality firm, the stranger-network generates greater sales than the friend-network when the network size is sufficiently large or the private signal is sufficiently accurate. We also examine the existence of experts and firms using advertising to influence consumers. Finally, we show how networks that are highly homogeneous or heterogeneous could impede observational learning.

No abstract

No abstract

Objective Prior studies have documented disparities in short and long sleep duration, excessive daytime sleepiness, and insomnia by educational attainment and race/ethnicity separately. We examined both independent and interactive effects of these factors with a broader range of sleep indicators in a racially/ethnically diverse sample. Methods We analyzed 2012 National Health Interview Survey data from 33,865 adults aged ≥18 years. Sleep-related symptomatology included short sleep duration (≤6 h), long sleep duration (≥9 h), fatigue >3 days, excessive daytime sleepiness, and insomnia. Bivariate analyses with chi-square tests and log-linear regression were performed. Results The overall age-adjusted prevalence was 29.1% for short sleep duration, 8.5% for long sleep duration, 15.1% for fatigue, 12.6% for excessive daytime sleepiness, and 18.8% for insomnia. Educational attainment and race/ethnicity were independently related to the five sleep-related symptoms. Among Whites, the likelihood of most sleep indicators increased as educational attainment decreased; relationships varied for the other racial/ethnic groups. For short sleep duration, the educational attainment-by-race/ethnicity interaction effect was significant for African Americans (p < 0.0001), Hispanics (p < 0.0001), and Asians (p = 0.0233) compared to Whites. For long sleep duration, the interaction was significant for Hispanics only (p = 0.0003). Conclusions Our results demonstrate the importance of examining both educational attainment and race/ethnicity simultaneously to more fully understand disparities in sleep health. Increased understanding of the mechanisms linking sociodemographic factors to sleep health is needed to determine whether policies and programs to increase educational attainment may also reduce these disparities within an increasingly diverse population.

Objectives To assess the association between sleep duration and health-related quality of life (HRQOL) among adults with or without chronic conditions. Methods Using the 2014 Behavioral Risk Factor Surveillance System, we analyzed self-reported data from adult respondents aged ≥18 years with (n = 277,757, unhealthy group) and without (n = 172,052. healthy group) reported history of any of nine chronic conditions (coronary heart disease, stroke, cancer, chronic obstructive pulmonary disease, diabetes, asthma, arthritis, depression, chronic kidney disease). Multivariable logistic regressions were separately constructed to assess the associations between sleep duration and four self-reported HRQOL measures after adjustment for sociodemographics, leisure-time physical activity, body mass index, and smoking status among unhealthy and healthy adults. Results The prevalence of poor/fair health, frequent physical distress, frequent mental distress, frequent activity limitation, and short sleep duration was 27.9%, 19.3%, 17.0%, 13.6%, and 38.3% in the unhealthy group and 6.9%, 4.0%, 5.3%, 2.1%, and 31.0% in the healthy group, respectively. U-shaped relationships of sleep duration to all four HRQOL indicators were observed among the unhealthy group and to poor/fair health, frequent mental distress, and frequent activity limitation among the healthy group. The relationships further varied by sex, age, race/ethnicity, and BMI category among the healthy group. Conclusions Relationships between extreme sleep duration and HRQOLs were observed among both healthy and unhealthy groups. These results can help inform public awareness campaigns and physician-counseling regarding the importance of sleep for mental health and well-being.