Uncovering the Source of Evaluation Bias in Micro-Lending

2021 | International Conference on Information Systems | Citations: 0

Authors: Hu, Xiyang; Huang, Yan; Li, Beibei; Lu, Tian

Abstract: We develop a structural econometric model to capture the decision dynamics of h ... Expand

Abstract: We develop a structural econometric model to capture the decision dynamics of human evaluators on an online micro-lending platform, and estimate the model parameters using real-world data. We find two types of biases in gender, i.e. preference-based bias and belief-based bias, are present in human evaluators' decisions. Both types of biases are in favor of female applicants. Through counterfactual simulations, we quantify the effect of gender bias on loan granting outcomes and the welfare of the company and the borrowers. Our results imply that both the existence of the preference-based bias and that of the belief-based bias reduce the company's profits. When the preference-based bias is removed, the company earns more profits. When the belief-based bias is removed, the company's profits also increase. Both increases result from lowering the approval probability for borrowers, especially female borrowers, who eventually default on loans. For borrowers, the elimination of either bias decreases the gender gap in the credit risk evaluation. Collapse

Topics: decision making crowdsourcing global navigation satellite system electronic finance

Methods: simulation machine learning qualitativ observational study econometric model experiment

Design Challenge Levels in E-Learning? Insights from a Large-Scale Field Experiment

2020 | International Conference on Information Systems | Citations: 0

Authors: Li, Andy Tao; Liu, De; Xu, Sean Xin

Abstract: The existing adaptive e-learning literature focuses on detecting and mitigating ... Expand

Abstract: The existing adaptive e-learning literature focuses on detecting and mitigating knowledge gaps but has paid scant attention to the issue of regulating the challenge levels. The latter issue is especially relevant when it comes to sequence the set of practice problems. Insights from flow theory suggest that learners would be more engaged if their perceived challenges match their abilities. Despite these theoretical predictions, there are gaps in terms of how to administrate the desirable level of challenge in dynamic learning environments and whether it pays to do so despite the fact that one may not perfectly control the level of challenge. Field experiments at seven middle schools reveal that steady challenge benefits both weak and strong learners compared with fluctuating challenges. Moreover, the optimal challenge is heterogeneous on learner’s preparations, and weak learners benefit more from low challenges, while strong learners are relatively insensitive to the challenge level. Collapse

Topics: online learning system logistics management data visualization affordance usability

Methods: experiment experimental group field experiment computational algorithm ridge regression

Theories: flow theory attribution theory

Detecting Review Manipulation on Online Platforms with Hierarchical Supervised Learning

2018 | Journal of Management Information Systems | Citations: 102

Authors: Kumar, Naveen; Venugopal, Deepak; Qiu, Liangfei; Kumar, Subodha

Abstract: Opinion spammers exploit consumer trust by posting false or deceptive reviews th ... Expand

Abstract: Opinion spammers exploit consumer trust by posting false or deceptive reviews that may have a negative impact on both consumers and businesses. These dishonest posts are difficult to detect because of complex interactions between several user characteristics, such as review velocity, volume, and variety. We propose a novel hierarchical supervised-learning approach to increase the likelihood of detecting anomalies by analyzing several user features and then characterizing their collective behavior in a unified manner. Specifically, we model user characteristics and interactions among them as univariate and multivariate distributions. We then stack these distributions using several supervised-learning techniques, such as logistic regression, support vector machine, and k-nearest neighbors yielding robust meta-classifiers. We perform a detailed evaluation of methods and then develop empirical insights. This approach is of interest to online business platforms because it can help reduce false reviews and increase consumer confidence in the credibility of their online information. Our study contributes to the literature by incorporating distributional aspects of features in machine-learning techniques, which can improve the performance of fake reviewer detection on digital platforms. Collapse

Topics: online review email spam digital platform deception detection fake review

Methods: machine learning product review supervised learning logistic regression support vector machine

Interval-Valued Intuitionistic Fuzzy WASPAS Method: Application in Reservoir Flood Control Management Policy

2018 | Group Decision and Negotiation | Citations: 0

Authors: Mishra, Arunodaya Raj; Rani, Pratibha

Abstract: Interval-valued intuitionistic fuzzy sets (IVIFSs) are very flexible tool to cop ... Expand

Abstract: Interval-valued intuitionistic fuzzy sets (IVIFSs) are very flexible tool to cope with the uncertainty arises in multi-criteria decision making (MCDM) problems. In recent times, MCDM problems with interval-valued intuitionistic fuzzy information have achieved more attention from researchers in different areas and consequently, several MCDM methods have been extended for IVIFSs. In this paper, a novel approach based on WASPAS method is developed under IVIFSs. The developed method is based on the operators of IVIFSs, some amendments in the classical WASPAS method and a new process for calculation of criteria and decision experts’ weights. In process for calculating weights, new procedures is propoesd to compute the decision experts’ weights and criteria weights based on interval-valued intuitionistic fuzzy information measures (entropy, divergence and similarity measures) to achieve more realistic weights. Innovative information measures are developed based on the exponential function for IVIFSs to determine the weights of the criteria and decision experts. Since the uncertainty is an unavoidable feature of MCDM problems, the developed method can be a constructive tool for decision-making in an uncertain environment. Further, an uncertain decision making problem of reservoir flood control management policy is implemented with interval-valued intuitionistic fuzzy information, which reveals the effectiveness and reliability of the proposed IVIF-WASPAS method. To validate the result, comparative analysis with existing methods and sensitivity analysis are presented under interval-valued intuitionistic fuzzy environment. Collapse

Topics: decision making

Methods: trade-off analysis method VIKOR method sensitivity analysis descriptive statistic cross sectional research

Applying behavioral economics in predictive analytics for B2B churn: Findings from service quality data

2017 | Decision Support Systems | Citations: 6

Authors: Barfar, Arash; Padmanabhan, Balaji; Hevner, Alan

Abstract: Motivated by the long-standing debate on rationality in behavioral economics and ... Expand

Abstract: Motivated by the long-standing debate on rationality in behavioral economics and the potential of theory-driven predictive analytics, this paper examines the link between service quality and B2B churn. Using longitudinal B2B transactional data with service quality indicators provided by a large company, we present evidence that both rationality and bounded-rationality assumptions play significant roles in predicting organizational decisions on churn. Specifically, variables that relate to the assumed rationality of organizations appear to provide accurate predictions while, at the same time, variables that capture boundedly rational decision rules appear to play a role through “somatic states” that make organizations more sensitive to the rational variables. In addition to presenting a novel approach for predicting organizational decisions on churn, this paper offers theoretical and managerial insights as well as opportunities for future research at the intersection of behavioral economics and predictive analytics for decision-making. Collapse

Topics: business-to-business service quality extract transform load process database system decision making

Methods: machine learning longitudinal research time series analysis feature engineering sensitivity analysis

Theories: behavioral economics attribution theory theory of bounded rationality

Bridging Qualitative and Quantitative Methods in Organizational Research: Applications of Synthetic Control Methodology in the U.S. Automobile Industry

2016 | Organization Science | Citations: 0

Authors: Fremeth, Adam R.; Holburn, Guy L. F.; Richter, Brian K.

Abstract: We assess the utility of synthetic control, a recently developed empirical metho ... Expand

Abstract: We assess the utility of synthetic control, a recently developed empirical methodology, for applications in organizational research. Synthetic control acts as a bridge between qualitative and quantitative research methods by enabling researchers to estimate treatment effects in contexts with small samples or few occurrences of a phenomenon or treatment event. The method constructs a counterfactual of a focal firm, or other observational unit, based on an objectively weighted combination of a small number of comparable but untreated firms. By comparing the firm’s actual performance to its counterfactual replica without treatment, synthetic control estimates, under certain assumptions, the magnitude and direction of treatment effects. We illustrate and critique the method in the context of the U.S. auto industry by estimating (a) the effect of government intervention in Chrysler’s management from 2009 to 2011 on its sales volumes and (b) the impact of Toyota’s 2010 “acceleration crisis” on Camry sales. Collapse

Topics: organizational context organizational productivity decision making marketing management missing data

Methods: synthetic control method longitudinal research case study Student's t-test parametric test

A Foresight Support System Using MCDM Methods

2015 | Group Decision and Negotiation | Citations: 0

Authors: Ondrus, Jan; Bui, Tung; Pigneur, Yves

Abstract: In this paper, we demonstrate the design and use of a foresight support system ( ... Expand

Abstract: In this paper, we demonstrate the design and use of a foresight support system (FSS) combining two multi-criteria decision-making (MCDM) methods. Traditionally, foresight activities involves Delphi, focus group, or Estimate–Talk–Estimate techniques to collect opinions of an expert panel. Often, these techniques are not computerized and data visualization is rudimentary. Our highly-interactive FSS solves a number of inherent issues during the data collection, analysis, and results visualization processes. Despite that MCDM methods have been recommended for technology foresight, a validation with a real field experiment was still required. To evaluate our approach and FSS, we conducted a foresight exercise for the Swiss mobile payments market. Our research confirms that the use of MCDM methods supported with a computerized tool can enhance the foresight processes and results. Collapse

Topics: mobile payment mobile system Near Field Communication decision making system support

Methods: trade-off analysis method ELECTRE computational algorithm field experiment design science

selfsurvey.org: A Platform for Prediction-Based Benchmarking and Feedback-Enabled Survey Research

2015 | European Conference On Information Systems | Citations: 0

Authors: Winkler, Till J; Sarstedt, Marko; Keil, Marian; Rost, Paul

Abstract: This design research builds on the idea to combine the strengths of traditional ... Expand

Abstract: This design research builds on the idea to combine the strengths of traditional survey research with a more practice-oriented benchmarking approach. We present selfsurvey.org, an online survey platform that allows providing instant and respondent-specific feedback based on a scientifically grounded research model and a structural equation model-based prediction technique. Based on the partial least squares analysis results of a training dataset, selfsurvey employs a scoring algorithm to derive respondent-specific predicted scores, compares these with the observed scores, and provides visualized and text-based outputs. Our evaluation of selfsurvey in the context of a maturity benchmarking study provides an indication for the perceived usefulness of this artifact and its underlying scoring algorithm. We argue that this prediction-based approach, which goes far beyond the functionality of common univariate benchmarking tools, can be used for a wide range of survey studies and help increase the perceived relevance of academic survey studies to practice. Collapse

Topics: IT service management user experience IT manager IT workforce evaluation criteria

Methods: survey computational algorithm partial least squares regression structural equation modeling design science

Partial Least Squares Structural Equation Modeling Approach for Analyzing a Model with a Binary Indicator as an Endogenous Variable

2013 | Communications of the Association for Information Systems | Citations: 0

Authors: Bodoff, David; Ho, Shuk Ying

Abstract: In this paper, we focus on PLS-SEM’s ability to handle models with observable bi ... Expand

Abstract: In this paper, we focus on PLS-SEM’s ability to handle models with observable binary outcomes. We examine the different ways in which a binary outcome may appear in a model and distinguish those situations in which a binary outcome is indeed problematic versus those in which one can easily incorporate it into a PLS-SEM analysis. Explicating such details enables IS researchers to distinguish different situations rather than avoid PLS-SEM altogether whenever a binary indicator presents itself. In certain situations, one can adapt PLS-SEM to analyze structural models with a binary observable variable as the endogenous construct. Specifically, one runs the PLS-SEM first stage as is. Subsequently, one uses the output for the binary variable and latent variable antecedents from this analysis in a separate logistic regression or discriminant analysis to estimate path coefficients for just that part of the structural model. We also describe a method—regularized generalized canonical correlation analysis (RGCCA)—from statistics, which is similar to PLS-SEM but unequivocally allows binary outcomes. Collapse

Topics: technology adoption

Methods: partial least squares path modeling computational algorithm logistic regression structural equation modeling SEM tool

Short-term prediction models for server management in Internet-based contexts

2009 | Decision Support Systems | Citations: 0

Authors: Casolari, Sara; Colajanni, Michele

Abstract: Modern Internet applications run on top of complex system infrastructures where ... Expand

Abstract: Modern Internet applications run on top of complex system infrastructures where several runtime management algorithms have to guarantee high performance, scalability and availability. This paper aims to offer a support to runtime algorithms that must take decisions on the basis of historical and predicted load conditions of the internal system resources. We propose a new class of moving filtering techniques and of adaptive prediction models that are specifically designed to deal with runtime and short-term forecast of time series which originate from monitors of system resources of Internet-based servers. A large set of experiments confirm that the proposed models improve the prediction accuracy with respect to existing algorithms and they show stable results for different workload scenarios. Collapse

Topics: internet technology database system website MySQL usability

Methods: longitudinal research computational algorithm autoregressive integrated moving average experiment time series analysis