Search

In this paper we use historic score-reporting records and test-taker metadata to inform data-driven recommendations that support international students in their choice of undergraduate institutions for study in the United States. We investigate the use of Structural Topic Modeling (STM) as a context-aware, probabilistic recommendation method that uses test-takers' selections and metadata to model the latent space of college preferences. We present the model results from two perspectives: 1) to understand the impact of TOEFL score and test year on test-takers' preferences and choices and 2) to recommend to the test-taker additional undergraduate institutions for application consideration. We find that TOEFL scores can explain variance in the probability that test-takers belong to certain preference-groups and, by accounting for this, our system adjusts recommendations based on student score. We also find that the inclusion of year, while not significantly altering recommendations, does enable us to capture minor changes in the relative popularity of similar institutions. The performance of this model demonstrates the utility of this approach for providing students with personalized college recommendations and offers a useful baseline approach that can be extended with additional data sources. [ABSTRACT FROM AUTHOR]

One realm of AI, recommender systems have attracted significant research attention due to concerns about its devastating effects to society's most vulnerable and marginalised communities. Both media press and academic literature provide compelling evidence that AI-based recommendations help to perpetuate and exacerbate racial and gender biases. Yet, there is limited knowledge about the extent to which individuals might question AI-based recommendations when perceived as biased. To address this gap in knowledge, we investigate the effects of espoused national cultural values on AI questionability, by examining how individuals might question AI-based recommendations due to perceived racial or gender bias. Data collected from 387 survey respondents in the United States indicate that individuals with espoused national cultural values associated to collectivism, masculinity and uncertainty avoidance are more likely to question biased AI-based recommendations. This study advances understanding of how cultural values affect AI questionability due to perceived bias and it contributes to current academic discourse about the need to hold AI accountable. [ABSTRACT FROM AUTHOR]

When the uni-variate risk measure analysis is generalized into the multi-variate setting, many complex theoretical and applied problems arise, and therefore the mathematical models used for risk quantification usually present model risk. As a result, regulators have started to require that the internal models used by financial institutions are more precise. For this task, we propose a novel multi-variate risk measure, based on the notion of the Wasserstein barycenter. The proposed approach robustly characterizes the company's exposure, filtering the partial information available from individual sources into an aggregate risk measure, providing an easily computable estimation of the total risk incurred. The new approach allows effective computation of Wasserstein barycenter risk measures in any location–scatter family, including the Gaussian case. In such cases, the Wasserstein barycenter Value-at-Risk belongs to the same family, thus it is characterized just by its mean and deviation. It is important to highlight that the proposed risk measure is expressed in closed analytic forms which facilitate its use in day-to-day risk management. The performance of the new multi-variate risk measures is illustrated in United States market indices of high volatility during the global financial crisis (2008) and during the COVID-19 pandemic situation, showing that the proposed approach provides the best forecasts of risk measures not only for "normal periods", but also for periods of high volatility. [ABSTRACT FROM AUTHOR]

Online retailers use product ratings to signal quality and help consumers identify products for purchase. These ratings commonly take the form of either non-personalized, aggregate product ratings (i.e., the average rating a product received from a number of consumers such as "the average rating is 4.5/5 based on 100 reviews"), or personalized predicted preference ratings for a product (i.e., recommender-system-generated predictions for a consumer's rating of a product such as "we think you'd rate this product 4.5/5"). Ratings in either format can provide decision aid to the consumer, but the two formats convey different types of product quality information and operate with different psychological mechanisms. Prior research has indicated that each recommendation type can significantly affect consumer's post-experience preference ratings, constituting a judgmental bias, but has not compared the effects of these two common product-rating formats. Using a laboratory experiment, we show that aggregate ratings and personalized recommendations create similar biases on post-experience preference ratings when shown separately. Shown together, there is no cumulative increase in the effect. Instead, personalized recommendations tend to dominate. Our findings can help retailers determine how to use these different types of product ratings to most effectively serve their customers. Additionally, these results help to educate the consumer on how product-rating displays influence their stated preferences. [ABSTRACT FROM AUTHOR]

Objective: The rapid growth of online health social websites has captured a vast amount of healthcare information and made the information easy to access for health consumers. E-patients often use these social websites for informational and emotional support. However, health consumers could be easily overwhelmed by the overloaded information. Healthcare information searching can be very difficult for consumers, not to mention most of them are not skilled information searcher. In this work, we investigate the approaches for measuring user similarity in online health social websites. By recommending similar users to consumers, we can help them to seek informational and emotional support in a more efficient way.Methods: We propose to represent the healthcare social media data as a heterogeneous healthcare information network and introduce the local and global structural approaches for measuring user similarity in a heterogeneous network. We compare the proposed structural approaches with the content-based approach.Results: Experiments were conducted on a dataset collected from a popular online health social website, and the results showed that content-based approach performed better for inactive users, while structural approaches performed better for active users. Moreover, global structural approach outperformed local structural approach for all user groups. In addition, we conducted experiments on local and global structural approaches using different weight schemas for the edges in the network. Leverage performed the best for both local and global approaches. Finally, we integrated different approaches and demonstrated that hybrid method yielded better performance than the individual approach.Conclusion: The results indicate that content-based methods can effectively capture the similarity of inactive users who usually have focused interests, while structural methods can achieve better performance when rich structural information is available. Local structural approach only considers direct connections between nodes in the network, while global structural approach takes the indirect connections into account. Therefore, the global similarity approach can deal with sparse networks and capture the implicit similarity between two users. Different approaches may capture different aspects of the similarity relationship between two users. When we combine different methods together, we could achieve a better performance than using each individual method. [ABSTRACT FROM AUTHOR]

The US Department of Agriculture Forest Service's Collaborative Forest Landscape Restoration Program (CFLRP) was created in 2009 to fund collaborative, science-based restoration. The role of science in collaborative forest restoration efforts such as the CFLRP is evolving as decisionmaking is complicated by a multitude of social and ecological factors. However, empirical evidence about the use of science in this context is limited. This article addresses the following question: How do CFLRP collaboratives obtain scientific information and what role does science play? Interviews and a survey of five CFLRP collaboratives in the Western United States were used to assess science sources, communication, perceptions, and obstacles. The results indicate that CFLRP collaboratives access scientific information primarily from university-based organizations and federal agencies via oral modes of delivery. Although the role of science is still somewhat unclear, CFLRP collaboratives are committed to incorporating science, and effective science delivery approaches may help facilitate the use of science. Management and Policy Implications: The Collaborative Forest Landscape Restoration Program (CFLRP) explicitly calls for "collaborative, science-based restoration" and use of the "best available science." However, collaborative forest restoration is a complex process in which scientific outputs are some of many sources of information. This complexity makes access to usable scientific information critical if science is to play a significant role in planning, decisionmaking, and implementation. Understanding how CFLRP collaboratives access scientific information and view the role of science can improve our understanding of how to most effectively use science in this context. I found that the five CFLRP collaboratives surveyed in this article largely access scientific information from university-based sources and federal agencies, primarily the Forest Service. Meanwhile, the ecological sciences and silvicultural knowledge are the most common disciplines called on for scientific information. Science is primarily delivered in oral formats through in-person meetings and presentations, although notably, respondents preferred to receive scientific information in peer-reviewed articles. Although the role of science in CFLRP collaboratives is still somewhat unclear and contested, the participants in the CFLRP projects surveyed generally accept scientific information, and there is a clear commitment to incorporating science in decisionmaking. Therefore, I recommend that participants in CFLRP collaboratives and similar collaborative forest restoration efforts emphasize oral, in-person delivery of scientific information, as well as sustained interactions among scientists, collaborative stakeholders, decisionmakers, and implementers. Ongoing, face-to-face interactions can help to build trust and mutual respect, encourage transparency, and facilitate communication and development of a common language. [ABSTRACT FROM AUTHOR]

This research compares the fifty top-rated movies for 2000–2009 recommended by the Internet Movie Database (IMDb) with the holdings of ten medium-sized public libraries. The study shows that there was no relation between a movie's rank in the IMDb listing and the number of copies held. Relationships appeared between number of library holdings and country of production (U.S. productions favored), genre (family, sci-fi, animation, adventure favored), and Motion Picture Association of America (MPAA) rating (G, PG, PG-13 preferred). Library holdings seem to reflect preferences of IMDb raters aged 30 and older, not popular movie tastes. [ABSTRACT FROM PUBLISHER]

Purpose - This paper seeks to review current research on assistive consumer technologies (ACT 1.0) and to discuss a series of research challenges that need to be addressed before the field can move towards tools that are more effective and more readily adopted by consumers (ACT 2.0). Design/methodology/approach - This is a conceptual paper. The perspective, commensurate with the current research and areas of expertise, is that of consumer researchers. Findings - The paper argues that, while substantial advances have been made in the technical design of ACTs - and the algorithms that power recommendation systems, there are substantial barriers to wide-scale consumer adoption of such tools that need to be addressed. In particular, future ACT designs will need to better integrate current research in human judgment and decision making to improve the ease with which such tools can be used. Originality/value - From the perspective of consumer researchers, the paper highlights a set of key areas of enquiry that have the potential to substantially advance assistive consumer technology research. [ABSTRACT FROM AUTHOR]

The article focuses on the new recommendations of the U.S. Preventive Services Task Force (USPSTF) in screening women with lipid disorders. The new scheme of USPSRF suggests that women will only be screened for lipid disorders if the risk factors of coronary heart disease (CHD) are seen to them. The recommendations for the screening of men are still the same. However, the recommendation for women has conflicted with the proposition of the National Institutes of Health and the American Heart Association.

In this paper, we discuss an explicit model function that can estimate the total number of deaths in the population, and particularly, estimate the cumulative number of deaths in the United States due to the current Covid-19 virus. We compare the modeling results to two related existing models based on a new criteria and several existing criteria for model selection. The results show the proposed model fits significantly better than the other two related models based on the U.S. Covid-19 death data. We observe that the errors of the fitted data and the predicted data points on the total number of deaths in the U.S. on the last available data point and the next coming day are less than 0.5% and 2.0%, respectively. The results show very encouraging predictability for the model. The new model predicts that the maximum total number of deaths will be approximately 62,100 across the United States due to the Covid-19 virus, and with a 95% confidence that the expected total death toll will be between 60,951 and 63,249 deaths based on the data until 22 April, 2020. If there is a significant change in the coming days due to various testing strategies, social-distancing policies, the reopening of community strategies, or a stay-home policy, the predicted death tolls will definitely change. Future work can be explored further to apply the proposed model to global Covid-19 death data and to other applications, including human population mortality, the spread of disease, and different topics such as movie reviews in recommender systems. [ABSTRACT FROM AUTHOR]

• The wavelet soft threshold denoising (WSTD) is systematically introduced. • The problem that is hardly tackled by decomposition methods is solved via the WSTD. • A deep learning method gated recurrent unit (GRU) is successfully developed. • The WSTD-GRU has a good performance in accuracy, speed, volatility and adaptability. • The parameters of the proposed model are adjusted by cross-validated grid-search. Wind speed forecasting plays a pivotal role in the security and economy of the power system operation. However, accurate prediction on the wind speed value is still quite challenging due to three features of wind speed: randomness, fluctuation and unpredictability. Though many previous works have studied decomposition-based methods for preprocessing, these methods are difficult to use in actual prediction, since the newly acquired data will greatly affect the values of the initial decomposed subsequences. To solve this problem, a good solution referred as the wavelet soft threshold denoising (WSTD) is proposed herein, which is also systematically introduced for the first time. Moreover, a novel deep learning method gated recurrent unit (GRU) is successfully developed to be combined with WSTD to forecast the wind speed series in this paper. In the ensemble WSTD-GRU model, the WSTD is applied to filter the redundant information from the original data of wind speed series; GRU, which is particularly advantageous for learning the features in case of cooperating with WSTD, is adopted to predict the future multi-step wind speed values. Also, the parameters of the GRU are adjusted by cross-validated grid-search, and they are visualized in the paper. In experiments, thirty-five commonly used models are involved for comparison to evaluate the forecasting performance of the proposed model, through the datasets from four sites in the United States covering a wide range of longitudes. Finally, a superior performance of this methodology with relatively high accuracy, fast forecasting speed, small volatility of results and good adaptability is verified by the test results of case studies aforementioned. [ABSTRACT FROM AUTHOR]