International Journal of Information Technology & Decision Making

The novel coronavirus disease — COVID-19 is a historic catastrophe that has caused many devastating impacts on human life and wellness. Researchers in academia and industry strive to understand the causes of this pandemic disease and find new therapeutics combating it. Consequently, the number of COVID-19 related publications increases rapidly, and it is too difficult for medical researchers and practitioners to keep up with the latest research and development. Literature filtering and categorization, and knowledge discovery can use text mining as a powerful tool. In this paper, we propose a text mining method to explore the categories of COVID-19 related themes and identify the standard methodologies that have been used. We discuss the potential limitations of this preliminary study and present future perspectives related to COVID-19 research. This paper provides an quantitative and qualitative mixed analysis example of using some research papers by data mining method to dig out several hidden information and set up a foundation for data scientists to develop more effective algorithms to deal with COVID-19 related problems.

Recently, the Internet of Things (IoT) has become a buzzword in various technology fields because of its many applications. Healthcare is one of the most important fields in daily life and holds significant interest for IoT and artificial intelligence researchers. In the area of healthcare, the mental healthcare field has attracted many researchers and funding organizations for various reasons. Among those reasons are the adverse impact of mental conditions on both individuals and society, the high costs of mental care, the nature of mental conditions and their hidden and unclear symptoms, and the stigma associated with them. In this work, an approach to building an IoT mental health monitoring and detection system has been proposed. The depression problem is used as a case study for the mental conditions. The proposed approach involves data collection, data pre-processing, feature extraction, feature selection, and model building. Machine learning (ML) and data mining techniques, along with different data types for training and testing the models, have been utilized to accomplish the tasks of the proposed approach. Four ML algorithms have been investigated to detect and predict the diagnosis of the mental state. These algorithms are random forest, decision tree, support vector machine (SVM) and k-nearest neighbors (KNN). Our ML detection model has achieved a high detection accuracy of 95% using the random forest algorithm. Comparisons in different aspects with other works are presented. Our work has the advantages of using ML, adopting different data types, and therefore achieving better classification.

We propose a hybrid approach for solving problems of mixed-integer linear programming with two types of variables (MILPTTV). The hybrid approach includes three algorithms: (1) Benders decomposition; (2) the ant colony algorithm; (3) the feasible heuristic method. We decompose the programming problem into a master problem and a subproblem using Benders decomposition. The ant colony algorithm generates an initial solution for the master problem. A feasible heuristic is used to obtain feasible solutions as input to the subproblem, whereas the subproblem is solved to optimality by using a linear programming solver that is named Linprog function in MATLAB R2016a. Over successive iterations, the master problem is refined by adding cuts from the subproblem and the master problem is solved by a linear programming solver Cplexmilp function. We compare the performance of the proposed hybrid approach against a general Benders decomposition approach as well as against a mixed-integer programming solver CPLEX (version 12.6) that is invoked by MATLAB on MIPLIB test problem mas76, mas74 and some large-scale numerical examples.

This research proposes a framework to improve the efficiency of classification and matching of descriptions of skill on resumes with jobs vacancies using labor market intelligence over a dataset of resumes harvested from social networks. To carry out the experiments, a Kaggle dataset was downloaded containing information from the LinkedIn social network with more than 200,000 records that were later filtered and pre-processed to generate a topic model to classify the entire dataset. Later, using machine learning algorithms, prediction exercises were performed to determine the most efficient match. This model offers high percentages of efficiency when predicting the job position of a candidate of information technology (IT) areas This prediction is achieved due the reduction of categories in these areas generated by the creation of the corresponding topic model to match the resume with the job position.

To address the ambiguity, uncertainty, and imprecision inherent in failure mode and effect analysis (FMEA) assessments, we propose a probabilistic interval-value term set (PIVTS) to express decision-makers’ evaluations. This approach allows decision makers to express their quantitative assessments using interval-value terms with probabilities, providing a more comprehensive and expressive representation of people’s preferences than exact numbers. In a probabilistic interval-valued environment, we propose an enhanced failure mode ranking method based on elimination and choice translating reality-III (ELECTRE III). Our ranking method includes a comprehensive computing framework for PIVTS to assess the importance of risk factors. Additionally, we provide rules for addition, subtraction, and division when comparing the disparity between two PIVTSs. Based on these rules, we also introduce the concordance index, discordance index, and outranking degrees. To demonstrate the effectiveness of our approach, we provide an example of its application to the steel production process. The advantages of PIVTS as a judgment technique and ELECTRE III as a decision method make our approach a better alternative for addressing complex uncertainties and determining the risk ranking of failure modes.

Mesenchymal stem cell (MSC) transfusion has shown promising results in treating COVID-19 cases despite the limited availability of these MSCs. The task of prioritizing COVID-19 patients for MSC transfusion based on multiple criteria is considered a multi-attribute decision-analysis (MADA) problem. Although literature reviews have assessed the prioritization of COVID-19 patients for MSCs, issues arising from imprecise, unclear and ambiguous information remain unresolved. Compared with the existing MADA methods, the robustness of the fuzzy decision by opinion score method (FDOSM) and fuzzy-weighted zero inconsistency (FWZIC) is proven. This study adopts and integrates FDOSM and FWZIC in a homogeneous Fermatean fuzzy environment for criterion weighting followed by the prioritization of the most eligible COVID-19 patients for MSC transfusion. The research methodology had two phases. The decision matrices of three COVID-19 emergency levels (moderate, severe, and critical) were adopted based on an augmented dataset of 60 patients and discussed in the first phase. The second phase was divided into two subsections. The first section developed Fermatean FWZIC (F-FWZIC) to weigh criteria across each emergency level of COVID-19 patients. These weights were fed to the second section on adopting Fermatean FDOSM (F-FDOSM) for the purpose of prioritizing COVID-19 patients who are the most eligible to receive MSCs. Three methods were used in evaluating the proposed works, and the results included systematic ranking, sensitivity analysis, and benchmarking checklist.

Due to the expansion of privacy protection, and data sharing gets complicated and is known to be a hot issue in the research areas. Many business scenarios in the supply chain network have been highly important in research areas. Especially, blockchain technology ensures the essential solution in the supply chain network for securing information sharing. Yet, it is difficult to maintain security at every blockchain level, whereas it is suggested to use “public–private key cryptography.” The fundamental goal of this research work is to present new privacy preservation of data using hybrid meta-heuristic algorithms with blockchain technology. The secured supply chain network is built with blockchain technology. The privacy preservation of data sharing in blockchain technology is handled by the electric fish-Harris hawks optimization (EF-HHO). This new algorithm is utilized for the key generation phase of two processes like “data sanitization and restoration.” The optimal key generation follows a multi-objective problem is solved by the variables like “Hiding Failure (HF) rate, information preservation (IP) rate, and false rule (FR) generation, and degree of modification (DM)”. The result analysis shows the offered model provides effective performance when compared with existing conventional techniques.

In this paper, we have established Cosine Trigonometric Operational Law (CTOL) and constructed Cosine Trigonometric Neutrosophic Weighted Averaging (CT-NWA) operator in the domain of generalized neutrosophic environment, which has been endorsed in a Multi-Criteria Group Decision Making (MCGDM) technique to select the best Anti-Pegasus software. In this study, the expertize recommendations and public opinion are well taken in context of debatable Pegasus issue in the Indian political, bureaucratic and democratic aspects. The problem has been addressed and resolved by using the integrated methodology of Analytical Network Process (ANP) and Evaluation based on Distance from Average Solution (EDAS) strategies. Here, we have optimized the weight factors and the gradation values of the decision makers as well as the sub-criteria of each criterion using Linear Programming (LP) model. Finally, we have performed the comparative analysis from various aspects to justify the reliability of our results.

Smart cities are becoming increasingly popular today due to various digital technologies in their infrastructure. One of the most critical challenges in the proper implementation and establishment of these cities is the violation of the security and privacy of users, which leads to citizens’ distrust and pessimism about smart city services. Many factors can affect the protection of users’ privacy. Identifying these factors helps developers understand what dimensions and aspects affect users’ privacy, through which they can provide appropriate solutions to protect the privacy of the user. Therefore, this study identifies and evaluates these dimensions. The authors proposed a novel methodology for identifying factors influencing user privacy in smart cities using the meta-synthesis method and the fuzzy decision-making trial and evaluation laboratory (DEMATEL). As a result of the meta-synthesis application, the authors identified seven dimensions of interest, specifically, Awareness and understanding, Trust, Self-control and personalization, Transparency, Reference standards frameworks and rules, Data management, and Security requirements. Moreover, using the DEMATEL technique and expert-based reviews, those seven dimensions were classified as cause and effect. As a result, the authors discovered that security requirements (D7) are the only cause dimension. As effect dimensions, “awareness and understanding (D1),” “data management (D6),” “trust (D2),” “frameworks and standards (D5),” “personalization and self-control (D3),” and “transparency (D4)” have been identified. Furthermore, the authors discovered that transparency is the most effective dimension in terms of affecting user privacy in smart cities. These findings may serve as guidelines for implementing the smart cities initiatives and protecting users’ privacy. This study also provides a discussion and insights for smart city developers.

Owing to the rapid growth of the Internet of Things (IoT) and mobile devices, big data has been generated, and a cloud computing system is required to manipulate it. The main issue in this system is the access delay (AD) when the IoT and mobile devices offload their workload on the servers and the workload balance among them. Mobile edge computing (MEC), which has a sufficient distribution of edge servers (ESs) and resources, has been developed to overcome this problem. The placement of ESs in MEC is an NP-hard problem and a multi-objective optimization (MOO) problem that aims to reduce the AD between mobile users and ES and balance the workload among these servers. In this paper, we propose a new hybrid natural-inspired algorithm by combining the manta ray foraging algorithm and the uniform crossover operator to find the optimal edge server placement (ESP) in MEC and minimize the AD between mobile users and the ES. Although the standard manta ray foraging optimization (MRFO) algorithm can balance diversification and intensification, it suffers from premature convergence. We invoke a uniform crossover operator in the standard MRFO algorithm to overcome this problem. In addition, the standard MRFO algorithm is designed to solve continuous optimization problems and is unsuitable for solving discrete problems. However, ESP is a discrete optimization problem. We modified the operators in the standard MRFO algorithm to solve the ESP problem, achieving an optimal solution in terms of minimizing the AD between BSs and ESs and balancing the workload among ESs. The proposed algorithm is then tested on the Shanghai Telecom dataset and compared with five other algorithms. The results show that the proposed algorithm achieves the least AD, and the workload balance reaches about 52% compared with the other algorithms.