1393 research outputs found
Sort by
An LSTM Based Approach for the Classification of Customer Reviews: An Exploratory Study
Get PDFSignificant research has been conducted to address the problem of identification and elimination of malicious content. The credibility of such information is always in question, especially in the E-commerce domain. This research proposes a classification model that automatically classifies customer reviews as credible or non-credible. This model encompasses a Long Short-Term Memory (LSTM) as a classification technique. The preliminary results have shown the potential of our model to classify customer reviews as credible/non-credible based on textual features
Prediction of Patient Willingness to Recommend Hospital: A Machine Learning-Based Exploratory Study
Get PDFHealth organizations are diligently working to achieve the zenith in service outcome and furtherance of patient satisfaction by embracing patient-centric policies. Patient recommendation is a critical indicator of patient satisfaction and hospital service quality. Evidence suggests that patient recommendation is the most valuable form of marketing. However, hospitals often encounter patients\u27 unwillingness to recommend them. Prior studies mainly rely on patient survey data to determine factors that impact patients’ willingness to recommend hospitals. Our study aims to identify factors that are not readily available in the patient surveys but have significant impact on hospital recommendation. Our proposed Machine Learning (ML) based model has incorporated multidimensional approach by identifying various affecting factors related to diverse hospital services for predicting the patient willingness to recommend the hospital. These factors will help providers to ameliorate quality of their services and implement more proactive measures that elevate hospital recommendations. Our results have shown that Random Forest (RF) to be the best technique for the prediction of hospital recommendation with a 0.08 RMSE and 0.59 adjusted R2. We have found that ED throughput, preventive care, and patient satisfaction related factors play a crucial role in influencing the patient\u27s decision to recommend the hospital
Credibility Analysis of Customer Reviews on Amazon: A Design Science Approach
Get PDFThis research examines the problem of identification and elimination of malicious customer reviews on Amazon.com. Online customer reviews are increasingly considered crowd-sourced consumer opinions that significantly influence online purchasing decisions (Hu, 2012). However, most current approaches to detecting fake reviews rely on either manual assessment of the reviews or the use of the mechanical Amazon Turks service (Mukherjee, 2014; Munzel, 2015). Manual assessment of customer reviews is not scalable in practice, leaving the quality of the current approaches to detect fake reviews questionable. The primary goal of our research is to develop a model of credibility analysis that automatically classifies amazon customer reviews as credible or non-credible. This model is developed based on the Design Science Research Methodology (Peffers, 2007) and encompasses a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM) as a classification technique. We first identify features of online customer reviews that can be used to effectively separate credible reviews from non-credible ones. Then fed the review dataset based on identified features to our proposed model for the assessment of review’s credibility. The study of existing literature indicates that most of current research on fake review focusses on the content of the reviews (Hu et al., 2012; Munzel, 2015).We, however, believe that content is only part of an effective method for detecting fake reviews. Our proposed model considers not only the textual but also the writing style and user related features of reviews. Further, we will compare our LSTM based model with other algorithms used in detecting misleading information such as Dynamic Series-Time Structure-based Support Vector machine (SVM-DSTS) (Ma, Gao, 2015) and Decision tree ranking (DTRank) (Zhao, 2015). Initial Design of proposed model will be presented for this TREO talk and encourage discussion concerning misleading customer reviews, existing fake review elimination initiatives, and Design science as an approach
Thucydides in Pyongyang: Fear, Honor and Interests in the 1968 Pueblo Incident
Get PDFPurpose: On January 23, 1968, North Korean naval forces captured a U.S spy ship, the USS Pueblo, off the coast of Wonsan. This incident nearly led to a second Korean War and heightened hostilities between the U.S and North Korean governments. This article demystifies the strategic thinking of Kim Il Sung’s regime and clarifies the reasoning behind Pyongyang’s risky undertaking in capturing the Pueblo and its crewmen as a rational and pragmatic action.
Design, Methodology, Approach: While the Pueblo crisis has been examined by a number of historians, this article which is based on former Eastern bloc archival documents and North Korean periodicals uses a multi-causal theoretical framework from an ancient Greek historian, Thucydides, in order to analyze the importance of fear, honor, and interest within North Korea’s military culture.
Findings: This article argues that North Korean regime’s fear of South Korea’s imminent economic supremacy and rising Japanese militarism along with defending the honor of Kim Il Sung and the DPRK’s territorial boundaries and advancing the interests of the global revolutionary movement factored greatly into Pyongyang’s decision-making process in 1968.
Practical Implications: In analyzing North Korea’s seemingly irrational aggression, it is vital to take a multi-causal approach, such as the one provided by Thucydides, into understanding North Korea’s past and present actions.
Originality, Value: Rather than arguing the 1968 Pueblo crisis as one motivated by internal or external concerns, this article posits that the North Korean leadership took a number of concerns into account and acted rationally in their capture of the Pueblo spy ship
Disentangling the effects of efficacy-facilitating informational support on health resilience in online health communities based on phrase-level text analysis
Get PDFThis study examines the different types of supportive messages posted on a forum at online Healthcare communities (OHCs), which facilitate user self-efficacy and response-efficacy and an issue of how such informational messages encourage users to enhance their health resilience via goal-setting for health improvement. We theorize that self-efficacy-oriented messages affect helpfulness, focusing on the efficiency of the implementation, while response-efficacy-oriented messages influence the relationships among helpfulness, goal-settings, and health resilience based on the outcome expectancy. Using a computer assisted approach which allows for the directed content analysis, we test a conceptual model with the text-data collected from an OHC
A Meta-Analysis of Evolution of Deep Learning Research in Medical Image Analysis
Get PDFWith a text mining and bibliometrics approach, we review the literature on the evolution of deep learning in medical image literature from 2012 – 2020 in order to understand the current state of the research and to identify the major research themes in image analysis to answer our research questions: RQ1: What are the learning modes that are evident in the literature? RQ2: What are the emerging learning modes in the literature? RQ3: What are the major themes in medical imaging literature? The analysis of 8704 resulting from a data collection process from peer-reviewed databases, our analysis discovered the six major themes of image segmentation studies, studies with image classification, evaluation procedures such as sensitivity and specificity, optical coherence tomography studies, MRI imaging studies, and Chest imaging studies. Additionally, we assessed the number of articles published each year, the frequent keywords, the author networks, the trending topics, and connections to other topics. We discovered that segmenting and classifying the images are the most common tasks. Transfer learning is the most researched area and cancer is the highly targeted disease and Covid-19 is the most recent research tren
The Use of Embedded Interaction Mechanisms for Low-Level Analysis Tasks
Get PDFThe use of information visualization is a strategy to reduce information overload and cognitive efforts. Interaction mechanisms aid the exploration of data when it is not practical to display all data points in one visual display. This study reports the results of a pilot study. The purpose of the study is to determine what interactive mechanisms are used and how they support a task or set of tasks