Rating Data Research Articles

Characterizing the reputation of evaluators using vectors in the object feature space

Evaluation is a frequent occurrence in our daily lives, especially in online systems. Establishing ways to characterize the reputation of an evaluator is therefore becoming an important problem in online systems, and this has attracted much attention. Most existing evaluation methods use the rating information in user–object bipartite networks to evaluate the reputations of evaluators. These methods use a scalar quantity to represent the reputation of each user, which is applied to try to identify malicious ratings and spamming attacks. In this paper, we suggest that the reputation should be characterized by a vector, and propose a reputation–evaluation algorithm based on the classification of objects. More specifically, the objects are classified into several categories according to the community structure information obtained from a one-mode projection onto the objects of a bipartite network. The reputation vector of an evaluator then represents his/her reputation for the objects in different categories. The evaluator is also assigned an appropriate attribute according to his/her reputation vector using K-means clustering. The results from both artificial and real rating data have shown that the presented method is better than a correlation-based ranking method in terms of both accuracy and robustness.

Online Learning of Parameters for Modeling User Preference Based on Bayesian Network

By analyzing users’ behavior data for personalized services, most state-of-the-art methods for user preference modeling are often based on batch-mode machine learning algorithms, where all rating data are assumed to be available throughout the training process. However, data in the real world often arrives sequentially and user preference may change dynamically. The real-time characteristics of rating data make the algorithms for preference modeling challenging to suit real-world online applications. By the user preference model (UPM) based on Bayesian network with a latent variable (BNLV), uncertain relationships among relevant attributes of users, objects and ratings could be represented, in which user preference is represented by the latent variable. In this paper, we propose an online approach for parameter learning of UPM. Specifically, we first extend the classic Voting EM algorithm by using Bayesian estimation in terms of the situation with latent variables. Consequently, we propose the algorithm for learning parameters of UPM from few and sequentially-changing rating data to reflect the gradually changing preferences. Finally, we test the effectiveness of our proposed algorithm by conducting experiments on various datasets. Experimental results demonstrate the superiority of our method in various measurements.

Geographical POI recommendation for Internet of Things: A federated learning approach using matrix factorization

SummaryWith the popularity of Internet of Things (IoT), Point‐of‐Interest (POI) recommendation has become an important application for location‐based services (LBS). Meanwhile, there is an increasing requirement from IoT devices on the privacy of user sensitive data via wireless communications. In order to provide preferable POI recommendations while protecting user privacy of data communication in a distributed collaborative environment, this paper proposes a federated learning (FL) approach of geographical POI recommendation. The POI recommendation is formulated by an optimization problem of matrix factorization, and singular value decomposition (SVD) technique is applied for matrix decomposition. After proving the nonconvex property of the optimization problem, we further introduce stochastic gradient descent (SGD) into SVD and design an FL framework for solving the POI recommendation problem in a parallel manner. In our FL scheme, only calculated gradient information is uploaded from users to the FL server while all the users manage their rating and geographic preference data on their own devices for privacy protection during communications. Finally, real‐world dataset from large‐scale LBS enterprise is adopted for conducting extensive experiments, whose experimental results validate the efficacy of our approach.

Credit Rating Prediction Through Supply Chains: A Machine Learning Approach

As supply chain channels physical, financial, and information flows as well as associated risks, a firm's supply chain information should be helpful in understanding and predicting its credit risks. Credit ratings, as an approximate but important measure of corporate credit risks, have been widely used by investors, creditors, and supply chain partners in their decision‐making. This study studies the role of supply chain information in predicting companies’ credit ratings. Using firm‐level supplier–customer linkages and corporate credit rating data, we develop a machine learning framework with gradient boosted decision trees to examine whether and what supply chain features can significantly improve the prediction accuracy of credit ratings, and what types of supply chain links have higher information content that positively affects the predictability of the supply chain features. We construct a firm's supply chain variables from its supplier and customer portfolios. We show that incorporating supply chain features can improve prediction accuracy over the benchmark credit rating model using only the focal firm's features. Moreover, the informativeness of supply chain links in focal credit risk prediction depends on the focal firm's industry sector, the relationship strength of such links, and the switching costs. Finally, we develop a focal credit rating prediction model with a high accuracy level using supply chain factors solely, which can potentially be applied to predict credit risks of small‐ and medium‐sized enterprises (SMEs) and private firms with no public financial information, as long as their supply chain information is available.

Improved recommender systems by denoising ratings in highly sparse datasets through individual rating confidence

Collaborative filtering (CF) is the most successful approach of Recommender Systems that has been applied in a wide range of applications. In this approach, historical rating data is exploited to calculate similarity between pairs of users and predict user preference over unseen items. The similarity is calculated using users' global preferences estimated from their previous interactions with items. To this end, commonly rated items by a given pair of users are utilized to calculate the similarity. In a sparse dataset, there are limited co-rated items. In addition, inconsistent behaviours of users in rating items may add some noise to data, which makes the ratings more untrustworthy leading to less accuraterecommendations. In this manuscript, we introduce an individual ratings confidence measure (IRC) to calculate the confidence of a given rating to each item by the target user. IRC consists of 5 factors that help to have more information about the interest and boredom of users. User ratings are denoised by prioritizing extreme and high-confidence ratings, and finally, the denoised ratings are used to obtain similarity values. This approach leads to a more accurate neighbourhood selection and rating prediction. We show the superiority of the proposed method when compared with state-of-the-art recommender algorithms over some well-known benchmark datasets.

Chimpanzee Extraversion scores vary with epigenetic modification of dopamine receptor gene D2 (DRD2) and early rearing conditions

ABSTRACT Chimpanzees have consistent individual differences in behaviour, also referred to as personality. Similar to human personality structure, five dimensions are commonly found in chimpanzee studies that show evidence for convergent and predictive validity (Dominance, Openness, Extraversion, Agreeableness, and Reactivity/Undependability). These dimensions are to some extent heritable, indicating a genetic component that explains part of the variation in personality scores, but are also influenced by environmental factors, such as the early social rearing background of the individuals. In this study, we investigated the role of epigenetic modification of the dopamine receptor D2 gene (DRD2) as a potential mechanism underlying personality variation in 51 captive chimpanzees. We used previously collected personality trait rating data and determined levels of DRD2 CpG methylation in peripheral blood samples for these same individuals. Results showed that DRD2 methylation is most strongly associated with Extraversion, and that varying methylation levels at specific DRD2 sites are associated with changes in Extraversion in nursery-reared, but not mother-reared, individuals. These results highlight the role of dopaminergic signalling in chimpanzee personality, and indicate that environmental factors, such as social experiences early in life, can have long-lasting behavioural effects, potentially through modification of the epigenome. These findings add to the growing evidence demonstrating the importance of the experience-dependent methylome for the development of complex social traits like personality.

Jointly Modeling Rating Responses and Times with Fuzzy Numbers: An Application to Psychometric Data

In several research areas, ratings data and response times have been successfully used to unfold the stagewise process through which human raters provide their responses to questionnaires and social surveys. A limitation of the standard approach to analyze this type of data is that it requires the use of independent statistical models. Although this provides an effective way to simplify the data analysis, it could potentially involve difficulties with regard to statistical inference and interpretation. In this sense, a joint analysis could be more effective. In this research article, we describe a way to jointly analyze ratings and response times by means of fuzzy numbers. A probabilistic tree model framework has been adopted to fuzzify ratings data and four-parameters triangular fuzzy numbers have been used in order to integrate crisp responses and times. Finally, a real case study on psychometric data is discussed in order to illustrate the proposed methodology. Overall, we provide initial findings to the problem of using fuzzy numbers as abstract models for representing ratings data with additional information (i.e., response times). The results indicate that using fuzzy numbers leads to theoretically sound and more parsimonious data analysis methods, which limit some statistical issues that may occur with standard data analysis procedures.

Feasibility, Enjoyment, and Language Comprehension Impact of a Tablet- and GameFlow-Based Story-Listening Game for Kindergarteners: Methodological and Mixed Methods Study.

BackgroundEnjoyment plays a key role in the success and feasibility of serious gaming interventions. Unenjoyable games will not be played, and in the case of serious gaming, learning will not occur. Therefore, a so-called GameFlow model has been developed, which intends to guide (serious) game developers in the process of creating and evaluating enjoyment in digital (serious) games. Regarding language learning, a variety of serious games targeting specific language components exist in the market, albeit often without available assessments of enjoyment or feasibility.ObjectiveThis study evaluates the enjoyment and feasibility of a tablet-based, serious story-listening game for kindergarteners, developed based on the principles of the GameFlow model. This study also preliminarily explores the possibility of using the game to foster language comprehension.MethodsWithin the framework of a broader preventive reading intervention, 91 kindergarteners aged 5 years with a cognitive risk for dyslexia were asked to play the story game for 12 weeks, 6 days per week, either combined with a tablet-based phonics intervention or control games. The story game involved listening to and rating stories and responding to content-related questions. Game enjoyment was assessed through postintervention questionnaires, a GameFlow-based evaluation, and in-game story rating data. Feasibility was determined based on in-game general question response accuracy (QRA), reflecting the difficulty level, attrition rate, and final game exposure and training duration. Moreover, to investigate whether game enjoyment and difficulty influenced feasibility, final game exposure and training duration were predicted based on the in-game initial story ratings and initial QRA. Possible growth in language comprehension was explored by analyzing in-game QRA as a function of the game phase and baseline language skills.ResultsEventually, data from 82 participants were analyzed. The questionnaire and in-game data suggested an overall enjoyable game experience. However, the GameFlow-based evaluation implied room for game design improvement. The general QRA confirmed a well-adapted level of difficulty for the target sample. Moreover, despite the overall attrition rate of 39% (32/82), 90% (74/82) of the participants still completed 80% of the game, albeit with a large variation in training days. Higher initial QRA significantly increased game exposure (β=.35; P<.001), and lower initial story ratings significantly slackened the training duration (β=−0.16; P=.003). In-game QRA was positively predicted by game phase (β=1.44; P=.004), baseline listening comprehension (β=1.56; P=.002), and vocabulary (β=.16; P=.01), with larger QRA growth over game phases in children with lower baseline listening comprehension skills (β=−0.08; P=.04).ConclusionsGenerally, the story game seemed enjoyable and feasible. However, the GameFlow model evaluation and predictive relationships imply room for further game design improvements. Furthermore, our results cautiously suggest the potential of the game to foster language comprehension; however, future randomized controlled trials should further elucidate the impact on language comprehension.

MSE-optimal K-factor of the Elo rating system for round-robin tournament

Abstract The Elo rating system contains a coefficient called the K-factor which governs the amount of change to the updated ratings and is often determined by empirical or heuristic means. Theoretical studies on the K-factor have been sparse and not much is known about the pertinent factors that impact its appropriate values in applications. This paper has two main goals: to present a new formulation of the K-factor that is optimal with respect to the mean-squared-error (MSE) criterion in a round-robin tournament setting and to investigate the effects of the relevant variables, including the number of tournament participants n, on the optimal K-factor (based on the model-averaged MSE). It is found that n and the variability of the deviation between the true rating and the pre-tournament rating have a strong influence on the optimal K-factor. Comparisons between the MSE-optimal K-factor and the K-factors from Elo and from the US Chess Federation as a function of n are also provided. Although the results are applicable to other sports in similar settings, the study focuses on chess and makes use of the rating data and the K-factor values from the chess world.

Fitting mixture models for feeling and uncertainty for rating data analysis

In this article, we present the command cub, which fits ordinal rating data using combination of uniform and binomial (CUB) models, a class of finite mixture distributions accounting for both feeling and uncertainty of the response process. CUB identifies the components that define the mixture in the baseline model specification. We apply maximum likelihood methods to estimate feeling and uncertainty parameters, which are possibly explained in terms of covariates. An extension to inflated CUB models is discussed. We also present a subcommand, scattercub, for visualization of results. We then illustrate the use of cub using a case study on students’ satisfaction for the orientation services provided by the University of Naples Federico II in Italy.

Recommendation Model Based on Dynamic Interest Group Identification and Data Compensation

With the increasing network service content, innovative methods are required for developing optimized network service for e-commerce companies. Accordingly, this study focuses on designing a framework containing personalization, interest group identification, and recommendation mechanisms. The primary contribution of this paper is to propose a recommendation model based on data compensation and dynamic user interest grouping. First, to address the problem of sparse user rating data, homeostasis compensation is performed on native data to more realistically restore the preference relationship between users and items by introducing the advantages of generative adversarial network in learning data distribution and enhancing data samples. Second, to address the problem of user interest generalization, information entropy is introduced to measure the user interest feature space. In addition, the time window marking method is used to further quantify the users’ dynamic interest group around the users’ interest drift. Finally, considering tensor decomposition characteristics in data dimension transformation and data compression, a score prediction model based on the “user-item-interest group” tensor decomposition is constructed. Simultaneously, a time decay function is introduced in the construction of the tensor to dynamically fit the user behavior and further improve prediction accuracy. Experiments show that the proposed framework can effectively improve the recommendation accuracies resulting from both sparse scoring data and dynamic user interest division.

Incorporating textual reviews in the learning of latent factors for recommender systems

• We use textual reviews to support the ratings in the latent factor model. • A review is interpreted as a description of the user/item and a description of the surrounding elements. • A latent factor model is proposed to account for both interpretations. • Reviews are incorporated not only into objective function but also into the initialization. In the field of recommender systems, the latent factor model is one of the state-of-the-art ones thanks to its strengths in accuracy and scalability. Its core is to learn latent factors for the representation of users and items using rating data collected through surveys after the users have experienced the items. However, on e-commerce applications, besides ratings, users can write reviews for items. A review generally indicates a user’s experience with an item while a rating indicates his/her level of satisfaction with such an experience. Latent factors can be learned more accurately if supported by such reviews. This study is distinctive in interpreting a review as both a description of the user/item and a description of the surrounding elements affecting the user's experience with the item. It has proven to be more effective than those that only consider a review as a description of the user/item. Especially, the analysis of the experimental results shows that our model provides supportive recommendations for users with detailed reviews in spite of their few collected ratings.

ANALYSIS OF THE ANXIETY ABOUT COVID-19 PANDEMIC WITH BREAST MILK PRODUCTION IN BREASTFEEDING MOTHERS

&lt;pre&gt;&lt;strong&gt;Background:&lt;/strong&gt; The postpartum period is a process that begins after the birth of the baby until the organs return to the way they were before pregnancy. There are complex physical and psychological changes. One of the psychological disorders that often occurs after childbirth is anxiety. This anxiety disorder will also indirectly affect the release of breast milk so that it interferes with the process of breastfeeding. The purpose of this study was to determine the relationship between anxiety during the COVID-19 pandemic and the production of breast milk in breastfeeding mothers. &lt;/pre&gt;&lt;pre&gt;&lt;strong&gt;Methods:&lt;/strong&gt; The design of this study was cross-sectional analytic. This study was carried out at East Denpasar Health Center I with a sample size of 30 respondents who met the inclusion criteria. The sampling technique of this research is purposive sampling. The independent variable in this study is anxiety and the dependent variable is the breast milk production. Data were collected using the Zung Self Rating Anxiety Scale (ZSAS) questionnaire and data analysis was carried out in stages including univariate and bivariate (Rank Spearman). &lt;/pre&gt;&lt;pre&gt;&lt;strong&gt;Result:&lt;/strong&gt; The results showed that most (60%) of respondents did not experience anxiety and most (60%) of respondents released breast milk quickly. Bivariate results show that there is a significant relationship between anxiety during the COVID-19 pandemic and the production of breast milk in breastfeeding mothers (p = 0.0001) with a correlation coefficient of 0.738. &lt;/pre&gt;&lt;pre&gt;&lt;strong&gt;Conclusion:&lt;/strong&gt; There is a relationship between anxiety and breastmilk production. The lower mother's anxiety, the better the production of her milk.&lt;strong&gt;&lt;/strong&gt;&lt;/pre&gt;

Benchmarking the banking sector of Bangladesh: a comprehensive analysis of performance and efficiency

PurposeThe economic development of Bangladesh is heavily reliant on the banking industry, yet it faces numerous hurdles, including liquidity issues, capital shortages, non-performing loans, inefficiencies and so on. Therefore, this study investigated the performance and efficiency of scheduled banks (state-owned, private commercial, foreign commercial and specialized banks) operating in Bangladesh.Design/methodology/approachThe research was conducted using secondary data from annual reports of banks. The CAMELS rating system and Data Envelopment Analysis (DEA) methods were employed to measure the performance and efficiency of banks, respectively.FindingsIn the overall bank rankings, results revealed that foreign commercial Standard Chartered Bank and state-owned Sonali Bank Limited came in first and last position, respectively. Among the four categories of banks, foreign commercial banks were the best performer, while state-owned banks were the worst. Only two banks, i.e. Citibank NA and HSBC Bank, were scale efficient while the remaining banks were inefficient. In terms of performance and efficiency, state-owned and specialized banks were deemed wanting.Practical implicationsThis study proposes recommendations to the policymakers that could lead to more effective tactics for improving the banking industry's performance and efficiency.Originality/valueAs far as the authors are concerned, this study presents empirical evidence on the performance and efficiency of different types of banks and explores comparisons among them, which has never been done to this extent in the country before.

Offline contexts of online jobs: Platform drivers, decent work, and informality in Lagos, Nigeria

SummaryMotivationThe Sustainable Development Goals targets include decent work for all by 2030 but progress in sub‐Saharan Africa has been slow. Over the past five years, the platform work sector (e‐hailing platforms in particular) has expanded considerably on the continent, providing work opportunities to the growing urban populations. The quality of this work, however, is heavily contested.PurposeWe deepen and extend our understanding of the gig economy in sub‐Saharan Africa and assess its potential for creating decent work. Our unique contribution lies in looking at the gig economy in comparison to the local threshold (the informal economy) as opposed to wage employment.Methods and approachWe first review the literature on the gig economy, the informal sector, and decent work in low‐ and lower‐middle‐income countries. We then draw on a case study of e‐hailing in Lagos, Nigeria, to assess the sector’s potential to create decent work. We discuss our findings against the four pillars of the International Labour Organization’s Decent Work Agenda: productive employment, labour standards, social protection, and social dialogue.FindingsWe find that e‐hailing platforms may offer certain gains compared to regular taxiing in the informal sector but these are context‐ and platform‐specific. Examples include access to insurance, credit, and cash transfers. Gig work remains embedded in the informal economy: it is unregulated and does not abide by labour standards nor offer social protections. We also find evidence of well‐organized labour movements.Policy implicationsE‐hailing may be productive work, as platform intermediation facilitates outreach and matching, translating into increased wages. This happens with the added cost of longer working hours in the absence of labour standards. We propose exploring the role of digital records kept by platforms (e.g. income volume and regularity, or customer ratings data) to facilitate the progressive transition of transport workers towards formality. We also recommend fostering industry standards (such as fair work certification) to protect workers’ rights.

COVID-19 and credit risk: A long memory perspective

The COVID-19 pandemic shows significant impacts on credit risk, which is the key concern of corporate bond holders such as insurance companies. Credit risk, quantified by agency credit ratings and credit default swaps (CDS), usually exhibits long-range dependence (LRD) due to potential credit rating persistence. With rescaled range analysis and a novel affine forward intensity model embracing a flexible range of Hurst parameters, our studies on Moody's rating data and CDS prices reveal that default intensities have shifted from the long-range to the short-range dependence regime during the COVID-19 period, implying that the historical credit performance becomes much less relevant for credit prediction during the pandemic. This phenomenon contrasts sharply with previous financial-related crises. Specifically, both the 2008 subprime mortgage and the Eurozone crises did not experience such a great decline in the level of LRD in sovereign CDS. Our work also sheds light on the use of historical series in credit risk prediction for insurers' investment.

Neural Metric Factorization for Recommendation

All current recommendation algorithms, when modeling user–item interactions, basically use dot product. This dot product calculation is derived from matrix factorization. We argue that an inherent drawback of matrix factorization is that latent semantic vectors of users or items sometimes do not satisfy triangular inequalities, which may affect the performance of the recommendation. Recently, metric factorization was proposed to replace matrix factorization and has achieved some improvements in terms of recommendation accuracy. However, similar to matrix factorization, metric factorization still uses a simple, linear fashion. In this paper, we explore leveraging nonlinear deep neural networks to realize Euclidean distance interaction between users and items. We propose a generic Neural Metric Factorization Framework (NMetricF), which learns representations for users and items by incorporating Euclidean metric factorization into deep neural networks. Extensive experiments on six real-world datasets show that, compared to the previous recommendation algorithms based purely on rating data, NMetricF achieves the best performance.

Combining review-based collaborative filtering and matrix factorization: A solution to rating's sparsity problem

An important factor affecting the performance of collaborative filtering for recommendation systems is the sparsity of the rating matrix caused by insufficient rating data. Improving the recommendation model and introducing side information are two main research approaches to address the problem. We combine these two approaches and propose the Review-Based Matrix Factorization method in this paper. The method consists of two phases. The first phase is review-based collaborative filtering, where an item-topic rating matrix is constructed by the feature-level opinion mining of online review text. This rating matrix is used to derive item similarities, which can be used to infer unknown users' ratings of the items. The second phase consists of rating imputation, where we first fill some of the empty elements of the user-item rating matrix, then conduct matrix factorization to learn the latent user and item factors to generate recommendations. Experiments on two actual datasets show that our method improves the accuracy of recommendation compared with similar algorithms.

Backfitting for large scale crossed random effects regressions

Regression models with crossed random effect errors can be very expensive to compute. The cost of both generalized least squares and Gibbs sampling can easily grow as N3/2 (or worse) for N observations. Papaspiliopoulos, Roberts and Zanella (Biometrika 107 (2020) 25–40) present a collapsed Gibbs sampler that costs O(N), but under an extremely stringent sampling model. We propose a backfitting algorithm to compute a generalized least squares estimate and prove that it costs O(N). A critical part of the proof is in ensuring that the number of iterations required is O(1), which follows from keeping a certain matrix norm below 1−δ for some δ>0. Our conditions are greatly relaxed compared to those for the collapsed Gibbs sampler, though still strict. Empirically, the backfitting algorithm has a norm below 1−δ under conditions that are less strict than those in our assumptions. We illustrate the new algorithm on a ratings data set from Stitch Fix.

A Secure and Efficient Wireless Charging Scheme for Electric Vehicles in Vehicular Energy Networks

To address the limited driving range of electric vehicles (EVs) and promote EVs’ penetration, vehicular energy networks (VENs) have emerged and opened possibility to charge EVs in motion via dynamic wireless power transfer (DWPT) technology. However, security and efficiency concerns arise due to the untrusted operating environment and EVs’ selfish charging/discharging behaviors. Existing trust models rely on the personal recommendations from neighboring EVs to identify malicious entities in VENs, which may cause potential privacy breaches and data misuse for recommenders. Besides, it is challenging to optimally schedule EVs’ energy behaviors by considering complex interactions among three energy entities (i.e., energy nodes, charging EVs, and discharging EVs). To this end, by leveraging blockchain technology and game theory, this paper proposes a secure and efficient wireless charging scheme to address these issues in VENs. Firstly, a blockchain-based fine-grained access control mechanism with traceability and auditability is presented to enable EV users to fully control and audit their personal rating data usage during trust management by logging data activities and issuing access tokens into decentralized ledgers. In this manner, the privacy of recommenders can be preserved by fully controlling the access and usage of personal rating data. Furthermore, by introducing cooperative wireless energy transfer mode, a hierarchical game-based energy scheduling algorithm is developed to optimize the strategies of three energy parties tier by tier, while considering their cooperation and competition. Finally, extensive simulations are conducted, which demonstrate that the proposed scheme can effectively improve users’ utility and security of energy transmission for EVs, compared with existing representative approaches.