Business Inferences and Risk Modeling with Machine Learning; The case of Aviation Incidents

2023 | HICSS | Citations: 0

Authors: Cankaya, B.; Topuz, K.; Glassman, A.

Abstract: Machine learning becomes truly valuable only when decision-makers begin to depe ... Expand

Abstract: Machine learning becomes truly valuable only when decision-makers begin to depend on it to optimize decisions. Instilling trust in machine learning is critical for businesses in their efforts to interpret and get insights into data, and to make their analytical choices accessible and subject to accountability. In the field of aviation, the innovative application of machine learning and analytics can facilitate an understanding of the risk of accidents and other incidents. These occur infrequently, generally in an irregular, unpredictable manner, and cause significant disruptions, and hence, they are classified as "highimpact, low-probability" (HILP) events. Aviation incident reports are inspected by experts, but it is also important to have a comprehensive overview of incidents and their holistic effects. This study provides an interpretable machine-learning framework for predicting aircraft damage. In addition, it describes patterns of flight specifications detected through the use of a simulation tool and illuminates the underlying reasons for specific aviation accidents. As a result, we can predict the aircraft damage with 85% accuracy and 84% in-class accuracy. Most important, we simulate a combination of possible flight-type, aircraft-type, and pilot-expertise combinations to arrive at insights, and we recommend actions that can be taken by aviation stakeholders, such as airport managers, airlines, flight training companies, and aviation policy makers. In short, we combine predictive results with simulations to interpret findings and prescribe actions.A machine-learning (ML) algorithm was chosen for its fast performance, short run-time, and interpretability. The ML algorithms were compared and evaluated on a confusion matrix for their mean receiver operating characteristic curve (ROC), Collapse

Topics: decision making decision support system graphical user interface usability missing data

Methods: simulation machine learning multinomial logistic regression computational algorithm support vector machine

Understanding Emojis for Financial Sentiment Analysis

2023 | International Conference on Information Systems | Citations: 0

Authors: Chen, Siyi; Xing, Frank

Abstract: ... Expand

Abstract: Collapse

Topics: application programming interface social media website cryptocurrency Python

Methods: sentiment analysis machine learning BERT deep learning experiment

Trust in Online Ride-Sharing Transactions: Impacts of Heterogeneous Order Features

2023 | Journal of Management Information Systems | Citations: 0

Authors: Cheng, Xusen; Fu, Shixuan; Sun, Jianshan; Zuo, Meiyun; Meng, Xiangsong

Abstract: With the development of the sharing economy, online ride-sharing has become a pr ... Expand

Abstract: With the development of the sharing economy, online ride-sharing has become a primary form of commuting. Using secondary transaction data, this study investigates the associations between the heterogeneous features and mutual trust in sharing economy-driven online ride-sharing transactions. Based on an examination of 12,404 ride-sharing orders in Beijing, we propose a set of trust distribution maps using order location data to reveal heterogeneous spatial patterns of the relationship between online ride-sharing transactions and mutual trust. The results show that the historical order completion rate and order distance are positively associated with mutual trust in ride-sharing transactions, whereas order time and departure density negatively and significantly influence mutual trust. Furthermore, we use machine learning algorithms to predict trust. The implications for theory and practice and future research directions are discussed. Collapse

Topics: ride sharing platform sharing platform logistics management geographic information system online review

Methods: machine learning computational algorithm regression analysis method experiment exploratory data analysis

Theories: signaling theory

ARTIFICIAL INTELLIGENCE-AUGMENTED DECISION MAKING IN SUPPLY CHAIN MONITORING: AN ACTION DESIGN RESEARCH STUDY

2023 | European Conference On Information Systems | Citations: 0

Authors: Herath Pathirannehelage, Savindu; Jóhannsson, Jóhann; Shrestha, Yash; Krogh, Georg von

Abstract: Organizations are progressively adopting hybrid human-artificial intelligence ( ... Expand

Abstract: Organizations are progressively adopting hybrid human-artificial intelligence (AI) systems in decisionmaking processes, with human decisions being augmented by AI insights. Among the promising AI applications in supply chain monitoring (SCMo) are predictive maintenance systems that predict potential device failures and augment maintenance decisions, allowing for timely and efficient interventions. Despite the growing proliferation of such systems, prescriptive knowledge encompassing technical, business, and organizational aspects on how to design, develop, and deploy them in actual operational environments remains limited. To address this shortcoming of evidence-based design principles in practice, our action design research developed a predictive maintenance system that predicts SCMo device failures and augments the maintenance decisions about those devices. By doing so, we outline generalizable design principles to guide prospective predictive maintenance systems in SCMo. Collapse

Topics: artificial intelligence predictive maintenance organizational context decision making systems evaluation

Methods: machine learning design principle design artifact artificial neural network random forest classification

Theories: evidence-based design

Responsible Artificial Intelligence in Healthcare: Predicting and Preventing Insurance Claim Denials for Economic and Social Wellbeing

2023 | Information Systems Frontiers | Citations: 1

Authors: Johnson, Marina; Albizri, Abdullah; Harfouche, Antoine

Abstract: It is estimated that one out of seven health insurance claims is rejected in the ... Expand

Abstract: It is estimated that one out of seven health insurance claims is rejected in the US; hospitals across the country lose approximately $262 billion annually due to denied claims. This widespread problem causes huge cash-flow issues and overburdens patients. Thus, preventing claim denials before claims are submitted to insurers improves profitability, accelerates the revenue cycle, and supports patients’ wellbeing. This study utilizes Design Science Research (DSR) paradigm and develops a Responsible Artificial Intelligence (RAI) solution helping hospital administrators identify potentially denied claims. Guided by five principles, this framework utilizes six AI algorithms – classified as white-box and glass-box – and employs cross-validation to tune hyperparameters and determine the best model. The results show that a white-box algorithm (AdaBoost) model yields an AUC rate of 0.83, outperforming all other models. This research’s primary implications are to (1) help providers reduce operational costs and increase the efficiency of insurance claim processes (2) help patients focus on their recovery instead of dealing with appealing claims. Collapse

Topics: artificial intelligence ADA missing data partial dependence plot responsible artificial intelligence

Methods: machine learning computational algorithm experiment support vector machine design artifact

ROLEX: A Novel Method for Interpretable Machine Learning Using Robust Local Explanations

2023 | Management Information Systems Quarterly | Citations: 2

Authors: Kim, Buomsoo (Raymond); Srinivasan, Karthik; Kong, Sung Hye; Kim, Jung Hee; Shin, Chan Soo; Ram, Sudha

Abstract: Recent developments in big data technologies are revolutionizing the field of h ... Expand

Abstract: Recent developments in big data technologies are revolutionizing the field of healthcare predictive analytics (HPA), enabling researchers to explore challenging problems using complex prediction models. Nevertheless, healthcare practitioners are reluctant to adopt those models as they are less transparent and accountable due to their black-box structure. We believe that instance-level, or local, explanations enhance patient safety and foster trust by enabling patient-level interpretations and medical knowledge discovery. Therefore, we propose the RObust Local EXplanations (ROLEX) method to develop robust, instance-level explanations for HPA models in this study. ROLEX adapts state-of-the-art methods and ameliorates their shortcomings in explaining individual-level predictions made by black-box machine learning models. Our analysis with a large real-world dataset related to a prevalent medical condition called fragility fracture and two publicly available healthcare datasets reveals that ROLEX outperforms widely accepted benchmark methods in terms of local faithfulness of explanations. In addition, ROLEX is more robust since it does not rely on extensive hyperparameter tuning or heuristic algorithms. Explanations generated by ROLEX, along with the prototype user interface presented in this study, have the potential to promote personalized care and precision medicine by providing patient-level interpretations and novel insights. We discuss the theoretical implications of our study in healthcare, big data, and design science. Collapse

Topics: explainable artificial intelligence usability user interface local interpretable model agnostic explanation problem solving

Methods: machine learning design artifact computational algorithm longitudinal research ablation analysis

Theories: lemon market theory

Multi-Channel Convolutional Neural Network for the Identification of Eyewitness Tweets of Disaster

2023 | Information Systems Frontiers | Citations: 0

Authors: Kumar, Abhinav; Singh, Jyoti Prakash; Rana, Nripendra P.; Dwivedi, Yogesh K.

Abstract: During a disaster, a large number of disaster-related social media posts are wid ... Expand

Abstract: During a disaster, a large number of disaster-related social media posts are widely disseminated. Only a small percentage of disaster-related information is posted by eyewitnesses. The post of a disaster eyewitness offers an accurate depiction of the disaster. Therefore, the information posted by the eyewitness is preferred over the other source of information as it is more effective at helping organize rescue and relief operations and potentially saving lives. In this work, we propose a multi-channel convolutional neural network (MCNN) that uses three different word-embedding vectors together to classify disaster-related tweets into eyewitness, non-eyewitness, and don’t know classes. We compared the performance of the proposed multi-channel convolutional neural network with several attention-based deep-learning models and conventional machine learning-models such as recurrent neural network, gated recurrent unit, Long-Short-Term-Memory, convolutional neural network, logistic regression, support vector machine, and gradient boosting. The proposed multi-channel convolutional neural network achieved an F1-score of 0.84, 0.88, 0.84, and 0.86 with four disaster-related datasets of floods, earthquakes, hurricanes, and wildfires, respectively. The experimental results show that the training MCNN model with different word embedding together performs better than the conventional machine-learning models and several other deep-learning models. Collapse

Topics: Twitter earthquake hurricane crowdsourcing social media

Methods: machine learning deep learning convolutional neural network experiment GloVe

Profit vs. Equality? The Case of Financial Risk Assessment and A New Perspective on Alternative Data

2023 | Management Information Systems Quarterly | Citations: 0

Authors: Lu, Tian; Zhang, Yingjie; Li, Beibei

Abstract: The importance of pursuing financial inclusion to accelerate economic growth an ... Expand

Abstract: The importance of pursuing financial inclusion to accelerate economic growth and enhance financial sustainability has been well noted. However, studies have provided few actionable insights into how financial institutions can balance the potential socioeconomic trade-off between profitability and equality. One major challenge arises from a lack of understanding of the impacts of various types of market information available on financial equality beyond economic profitability. Another challenge lies in how the socioeconomic trade-off under a large set of counterfactual policies in a real-world setting can be evaluated. Our motivation for the present study was the emerging sources of digitized user-behavior data (i.e., "alternative data") stemming from the high penetration of mobile devices and internet access. Accordingly, we investigated how alternative data from smartphones and social media can help mitigate potential financial inequality while preserving business profitability in the context of financial credit risk assessment. We partnered with a leading microloan website to design a novel "meta" experiment that allowed us to simulate various real-world field experiments under an exhaustive set of counterfactual policies. Interestingly, we found that profiling user financial risk using smartphone activities is 1.3 times more effective in improving financial inclusion than using online social media information (23.05% better vs. 18.11%), and nearly 1.3 times more effective in improving business profitability (42% better vs. 33%). Surprisingly, we found that using consumers' online shopping activities for credit risk profiling can hurt financial inclusion. Furthermore, we investigated potential explanations for financial inclusion improvements. Our findings suggest that alternative data, especially users' smartphone activities, not only demonstrate higher ubiquity but also appear to be more orthogonal to conventional sensitive demographic attributes. This, in turn, can help mitigate statistical bias driven by the unobserved factors or underrepresentative training samples in machine-based risk assessment processes. Collapse

Topics: website social media smartphone microblog electronic commerce

Methods: experiment XGBoost machine learning experimental design k-nearest neighbor

Fake review detection system for online E-commerce platforms: A supervised general mixed probability approach

2023 | Decision Support Systems | Citations: 0

Authors: Luo, Jiwei; Luo, Jian; Nan, Guofang; Li, Dahui

Abstract: Online consumer reviews play an important role in helping consumers judge the qu ... Expand

Abstract: Online consumer reviews play an important role in helping consumers judge the quality and authenticity of products on e-commerce platforms. However, the constant presence of fake reviews on these platforms has significantly impacted the operation and development of e-commerce platforms. In this study, we develop a novel supervised probabilistic method to detect fake reviews by utilizing the difference in the distribution of nonfraudulent reviews and that of fake reviews. Specifically, we first derive the univariate distributions of several unique features (linguistic, behavioral, and interrelationship features). We then integrate these distributions into two mixed distributions according to their labels to represent the overall difference between non-fraudulent reviews and fake reviews. Next, we randomly generate synthetic review data points with different labels from the above mixed distributions. Finally, we train a Multilayer Perceptron model by using these synthetic review data to obtain a classifier. We conducted several experiments to test the model using several original real-world review datasets. Numerical results indicated that the proposed supervised method outperformed some wellknown sampling models and fake review detection methods, in terms of classification accuracy. Moreover, we extend the proposed method to handle the scenarios with small samples of raw review data. This study con­ tributes to the literature by exploiting the difference in the distribution of non-fraudulent reviews and that of fraudulent reviews, which can improve the accuracy of fake review detection for online platforms. Collapse

Topics: fake review electronic commerce big data decision support system user behavior

Methods: computational algorithm experiment machine learning semi-supervised learning artificial neural network

Designing Interactive Explainable AI Systems for Lay Users

2023 | International Conference on Information Systems | Citations: 0

Authors: Meza Martínez, Miguel; Mädche, Alexander

Abstract: Explainability considered a critical component of trustworthy artificial intell ... Expand

Abstract: Explainability considered a critical component of trustworthy artificial intelligence (AI) systems, has been proposed to address AI systems' lack of transparency by revealing the reasons behind their decisions to lay users. However, most explainability methods developed so far provide static explanations that limit the information conveyed to lay users resulting in an insufficient understanding of how AI systems make decisions. To address this challenge and support the efforts to improve the transparency of AI systems, we conducted a design science research project to design an interactive explainable artificial intelligence (XAI) system to help lay users understand AI systems' decisions. We relied on existing knowledge in the XAI literature to propose design principles and instantiate them in an initial prototype. We then conducted an evaluation of the prototype and interviews with lay users. Our research contributes design knowledge for interactive XAI systems and provides practical guidelines for practitioners. Collapse

Topics: explainable artificial intelligence artificial intelligence shapley additive explanation open source enterprise architecture

Methods: design artifact design principle qualitative interview design requirement design science

Theories: game theory