Rating-based Model Research Articles

Multi-aspect Graph Contrastive Learning for Review-enhanced Recommendation

Review-based recommender systems explore semantic aspects of users’ preferences by incorporating user-generated reviews into rating-based models. Recent works have demonstrated the potential of review information to improve the recommendation capacity. However, most existing studies rely on optimizing review-based representation learning part, thus failing to explicitly capture the fine-grained semantic aspects, and also ignoring the intrinsic correlation between ratings and reviews. To address these problems, we propose a multi-aspect graph contrastive learning framework, named MAGCL, with three distinctive designs: (i) a multi-aspect representation learning module, which projects semantic relations to different subspaces by decoupling review information, and then obtains high-order decoupled representations in each aspect via graph encoder. (ii) the contrastive learning module performs graph contrastive learning to capture the correlation between rating and review patterns, which utilize unlabeled data to generate self-supervised signals and, in turn, relieve the data sparsity problem of supervision signals. (iii) the multi-task learning module conducts joint training to learn high-order structure-aware yet self-discriminative node representations by combining recommendation task and self-supervised task, which helps alleviate the over-smoothing problem. Extensive experiments are conducted on four real-world review datasets and the results show the superiority of the proposed framework MAGCL compared with several state of the arts. We also provide further analysis on multi-aspect representations and graph contrastive learning to verify the advantage of proposed framework.

Detecting Product Review Spammers Using Principles of Big Data

The growing consumerism has led to the importance of online reviews on the Internet. Opinions voiced by these reviews are taken into consideration by many consumers for making financial decisions online. This has led to the development of opinion spamming for profitable motives or otherwise. This work has been done to tackle the challenge of identifying such spammers, but the scale of the real-world review systems demands this problem to be tackled as a big data challenge. So, an effort has been made to detect online review spammers using the principle of big data. In this article, a rating-based model has been studied under the light of large-scale datasets (more than 80 million reviews by 20 million reviewers) using the Hadoop and Spark frameworks. Scale effects have been identified and mitigated to provide better context to large review systems. An improved computational framework has been presented to compute the overall spamcity of reviewers using exponential smoothing. The value of the smoothing factor was set empirically. Finally, future directions have been discussed.

Benchmark test for calculated default probability values derived from the application of rating-based models

Subject. This article deals with the issues of validation of the consistency of rating-based model forecasts. Objectives. The article aims to provide developers and validators of rating-based models with a practical fundamental test for benchmarking study of the estimated default probability values obtained as a result of the application of models used in the rating system. Methods. For the study, I used the classical interval approach to testing of statistical hypotheses focused on the subject area of calibration of rating systems. Results. In addition to the generally accepted tests for the correspondence of the predicted probabilities of default of credit risk objects to the historically realized values, the article proposes a new statistical test that corrects the shortcomings of the generally accepted ones, focused on "diagnosing" the consistency of the implemented discrimination of objects by the rating model. Examples of recognizing the reasons for a negative test result and negative consequences for lending are given while maintaining the current settings of the rating model. In addition to the bias in the assessment of the total frequency of defaults in the loan portfolio, the proposed method makes it possible to objectively reveal the inadequacy of discrimination against borrowers with a calibrated rating model, diagnose the “disease” of the rating model. Conclusions and Relevance. The new practical benchmark test allows to reject the hypothesis about the consistency of assessing the probability of default by the rating model at a given level of confidence and available historical data. The test has the advantage of practical interpretability based on its results, it is possible to draw a conclusion about the direction of the model correction. The offered test can be used in the process of internal validation by the bank of its own rating models, which is required by the Bank of Russia for approaches based on internal ratings.

Crowdlending: mapping the core literature and research frontiers

Peer-to-peer (P2P) lending uses two-sided platforms to link borrowers with a crowd of lenders. Despite considerable diversity in crowdlending research, studies in this area typically focus on several common research topics, including information asymmetries, social capital, communication channels, and rating-based models. This young research field is still expanding. However, its importance has increased considerably since 2018. This rise in importance suggests that P2P lending may offer a promising new scientific research field. This paper presents a bibliometric study based on keyword co-occurrence, author and reference co-citations, and bibliographic coupling. The paper thus maps the key features of P2P lending research. Although many of the most cited papers are purely financial, some focus on behavioral finance. The trend in this field is toward innovative finance based on new technologies. The conclusions of this study provide valuable insight for researchers, managers, and policymakers to understand the current and future status of this field. The variables that affect new financial contexts and the strategies that promote technology-based financial environments must be investigated in the future.

Exploring user movie interest space: A deep learning based dynamic recommendation model

Exploring user interest behind massive user behaviors is essential for online recommendations. Although recommendation models have been proposed recently with great success, existing studies ignore not only the timeliness of online users’ behaviors in terms of their interest, but also the sequential characteristics of their behaviors. To overcome this limitation, we construct a User Movie Interest Space (UMIS) model based on the sequential ratings of users. We define three indexes to elucidate the features of the interest of users for UMIS, which describe different patterns of behaviors of users related to their interests. Based on UMIS we propose a deep learning model named Dynamic Interest Flow (DIF) to provide dynamic movie recommendations. The DIF model achieves intelligently multi-dimensional observations on a user’s interest space and to predict simultaneously a variety of their future interests. Experimental results indicate that DIF outperforms traditional rating-based models and other state-of-the-art deep learning models. Results also demonstrate that modeling a dynamic recommendation as a sequential prediction is supposed to obtain outstanding advantages.

Methodical Principles of Evaluation of Competitive Ability of Construction Industry and Real Estate Development Companies

Based on systematization of methods and criteria of evaluation of competitive ability the present authors have substantiated the comparative principle of competitive ability. Selection algorithm of methodological instrumentarium for analytical competitiveness assessment was developed; a modification of a rating-based model of evaluation of competitive ability with the application of calculation and expert based methods have been presented as consequence. Obtained results in terms of time series provide representativity of arrays of information and significantly increase reliability of assessment. The method was tested by evaluating the competitive ability of several large real estate development companies and that of the construction industry.

A Regime Switching Model for the Term Structure of Credit Risk Spreads

We consider a rating-based model for the term structure of credit risk spreads wherein the credit-worthiness of the issuer is represented as a finite-state continuous time Markov process. This approach entails a progressive drift in credit quality towards default. A model of the economy is presented featuring stochastic transition probabilities; credit instruments are valued via an ultra parabolic Hamilton-Jacobi system of equations discretized utilizing the method-of-lines finite difference method. Computations for a callable bond are presented demonstrating the efficiency of the method.

Quantifying and Explaining Implicit Public Guarantees for European Banks

This study provides estimation of public implicit guarantees over the period 1997 to 2012 using a rating-based model. The analysis focuses on a sample of 56 large listed European banks. It appears that the main element for determining the value of the public subsidy is the intrinsic strength of the bank. In addition, we bring evidence on the importance of the guarantor strength on the value of the implicit guarantee: a higher sovereign rating of a bank's home country leads to larger implicit subsidies for banks' debt. Our findings also suggest that the recently observed decrease in the value of implicit subsidies goes beyond the declining in European sovereigns' strength. Rather, it is consistent with the implementation of resolution regimes and practices moving from a bailout resolution policy to bail-in recapitalization. Resolution schemes providing more explicit rules to treat banks' bankruptcies will reduce investors' expectations of public support.

On the Pricing of Step-Up Bonds in the European Telecom Sector

This paper investigates the pricing of step-up bonds, i.e. corporate bonds with provisions stating that the coupon payments increase as the credit rating level of the issuer declines. To assess the risk-neutral rating transition probabilities necessary to price these bonds, we introduce a new calibration method within the reduced-form rating-based model of Jarrow, Lando, and Turnbull (1997). We also treat split ratings and adjust for rating outlook. Step-up bonds have been issued in large amounts in the European telecom sector, and we find that, through most of the sample, step-up bonds issued by the two largest issuers have traded at a discount relative to comparable fixed-coupon bonds from the same issuers. Our findings cannot be attributed to traditional liquidity factors, and they suggest that issuing step-up bonds increased the cost of capital for the issuers.

A cross-validity comparison of rating-based and choice-based conjoint analysis models

This paper compares OLS, hierarchical Bayes (HB), and latent segment, rating-based conjoint models to HB and latent segment choice-based conjoint models. The biggest systematic difference between rating- and choice-based parameters is consistent with a compatibility effect. This leads to a stronger prominence effect for rating-based models. The HB rating-based model has the highest hit rate and choice share validations and the OLS model has the second highest. Within both rating- and choice-based models, hierarchical Bayes models have higher hit rate and choice share validations than latent segment models. Using a choice simulator, the profiles predicted to be optimal with an HB choice-based model are similar to those predicted to be optimal with an HB rating-based model.

Swap Pricing with Two-Sided Default Risk in a Rating-Based Model

This paper analyzes the pricing of defaultable securities in rating-based models where the default of more than one agent is involved. We extend the model of Duffie and Huang (1996) to a framework which explicitly takes into account the rating of each party. Our extension allows us to investigate the effects on swap spreads of early termination provisions, i.e., credit triggers, which are linked to the ratings of contracting parties.

Comment on ‘Swap Pricing with Two-Sided Default Risk in a Rating-Based Model’

Comment on ‘Swap Pricing with Two-Sided Default Risk in a Rating-Based Model’

Swap Pricing with Two-Sided Default Risk in a Rating-Based Model *

Swap Pricing with Two-Sided Default Risk in a Rating-Based Model *