Ebook: Modern Management based on Big Data II and Machine Learning and Intelligent Systems III
It is data that guides the path of applications, and Big Data technologies are enabling new paths which can deal with information in a reasonable time to arrive at an approximate solution, rather than a more exact result in an unacceptably long time. This can be particularly important when dealing with an urgent issue such as that of the COVID-19 pandemic.
This book presents the proceedings of two conferences: MMBD 2021 and MLIS 2021.
The MMBD conference deals with two main subjects; those of Big Data and Modern Management. The MLIS conference aims to provide a platform for knowledge exchange of the most recent scientific and technological advances in the field of machine learning and intelligent systems. Both conferences were originally scheduled to be held from 8-11 November 2021, in Quanzhou, China and Xiamen, China respectively. Both conferences were ultimately held fully online on the same dates, hosted by Huaqiao University in Quanzhou and Xiamen respectively.
The book is in two parts, and contains a total of 78 papers (54 from MMBD2021 and 24 from MLIS2021) selected after rigorous review from a total of some 300 submissions. The reviewers bore in mind the breadth and depth of the research topics that fall within the scope of MMBD and MLIS, and selected the 78 most promising and FAIA mainstream-relevant contributions for inclusion in this two-part volume. All the papers present original ideas or results of general significance supported by clear reasoning, compelling evidence and rigorous methods.
Data guides the path of applications. Big data technologies are enabling new paths which can deal with information in a reasonable time and arrive at an approximate solution, rather than a more exact result in an unacceptably long time. This is particularly important today as we continue to work against the COVID-19 pandemic.
The conference series Modern Management based on Big Data (MMBD) has held its second edition, whereas the conference series Machine Learning and Intelligent Systems (MLIS) has held its third edition.
MMBD deals with two main branches; those of Big Data and Modern Management. The former includes: data capture and storage; search, sharing, and analytics; big data search, mining and visualization; big data technologies; data visualization; architectures for massive parallel processing; data mining tools and techniques; machine learning algorithms for big data; cloud computing platforms; distributed file systems and databases; scalable storage systems; and big data for business, government and society, among other things. The latter encompasses: modern management; management strategy; management decision making; manufacturing systems; logistic systems; facilities planning; cost analysis; engineering economy; total quality management; management information systems; human factor engineering; and human resources.
The MLIS conference is an annual conference that aims to provide a platform for knowledge exchange of the most recent scientific and technological advances in the field of machine learning and intelligent systems. It also aims to strengthen links in related fields within the scientific community.
This book, which is in two parts, contains accepted papers presented at MMBD 2021 and MLIS 2021. The MMBD conference was originally scheduled to be held from 8–11 November 2021 in Quanzhou, China and MLIS was originally scheduled to be hosted in Xiamen, China on the same dates. The most popular topics in this book concern data analysis and big data applications.
All papers have been conscientiously reviewed by programme committee members, who bore in mind the breadth and depth of the research topics that fall within the scope of MMBD and MLIS respectively. From the some 300 submissions received for MMBD and MLIS, the 78 most promising and FAIA mainstream-relevant contributions have been included in this two-part volume. These present original ideas or results of general significance supported by clear reasoning, compelling evidence and rigorous methods.
I would like to thank all the keynote and invited speakers, as well as the authors and anonymous reviewers, for their effort in making both MMBD and MLIS conferences of the highest standard. We would also like to take this opportunity to express our gratitude to all those people, especially the programme committee members and reviewers, who devoted time to assessing the papers. It is an honour to have been involved from the outset with the publication of these proceedings, which will form part of the prestigious series Frontiers in Artificial Intelligence and Applications (FAIA) from IOS Press. Our particular thanks also go to J. Breuker, N. Guarino, J.N. Kok, R. López de Mántaras, J. Liu, R. Mizoguchi, M. Musen, S.K. Pal and N. Zhong, the FAIA series editors, for supporting this conference.
Last but not least, any inconvenience caused due to the format change from face to face to virtual is sincerely regretted. Hopefully we will meet face to face at the MMBD2022 and MLIS 2022 conferences next year, the venues for which have not been decided at the time of publication.
Antonio J. Tallón-Ballesteros
University of Huelva (Spain)
Huelva city, Spain
Taking as a basis the elaboration likelihood model (ELM), this paper evaluates how the way in which the customer searches, evaluates and compares information influences the development of omni-channel behaviour, as well as each of its most common practices, webrooming (researching products online but purchasing products in a physical store) and showrooming (visiting physical stores to check out products and then buying them online). The results obtained from a sample of 939 apparel shoppers using the database constructed for the Spanish retail sector by GfK reflect that compared to one-stop shoppers, omni-shoppers (without distinguishing specific typologies) spend more time and effort planning their decision-making. The combination of physical and virtual channels makes it easier for the consumer to be more involved in the shopping experience and to search, compare and evaluate specific information about the product and/or retailer before the final purchase. This more reflective behaviour in which more time is spent on the consumer journey and more information is handled is what ELM defines as the central information processing route. While it is true that the central information processing route predominates, it is concluded when analysing each of the omni-channel behaviours separately that webroomers are more likely to follow this route, analysing in depth all issues related to the product they want to buy. On the other hand, although showrooming behaviour cannot be associated with the same intensity to the central information processing route, nor can it be associated with a less planned customer journey, like that of e-shoppers who focus only on prices and cost savings associated with the purchase. Showroomers use the internet to learn about retailer-related aspects as well as other consumers’ opinions of the product before buying the product from the online store. Taking these results into account, managers should keep in mind the idea that webroomers and showroomers are as different as they are the same. Thus, both the internet and the physical store have to serve as both an information point and a shopping channel. Websites need to be usable and simple so that webroomers can get in-depth information about the retailer’s portfolio and showroomers can make a purchase in a few quick steps. On the other hand, the physical store will be a touchpoint where omni-shoppers will enjoy unique experiences, highlighting the sales force that will be key for webroomers and showroomers to develop a stronger bond with the firm and not shop at any other competitor retailer.
With the advent of knowledge economy, any enterprise can be regarded as providing “knowledge service” for customers. Through the process of tacit knowledge acquisition and innovation, enterprises will eventually transform tacit knowledge into the core competitiveness of enterprises. Based on the sample data of Chinese listed manufacturing companies for three consecutive years, this paper analyzes the impact of three ways of tacit knowledge acquisition (namely: employee training, research and development and market research) on enterprise performance, and introduces the factor of redundant resources to analyze its moderating effect on this relationship It is found that the three ways of tacit knowledge acquisition have significant positive effects on enterprise performance; different types of redundant resources (non precipitation redundant resources, precipitation redundant resources) strengthen and weaken the three positive relationship mechanisms. Based on the theory of resource-based view, it promotes the development process of knowledge acquisition and innovation under the knowledge economy of enterprises.
Sentiment analysis has received much attention in Information Retrieval (IR) and other domains including data mining, machine learning algorithms and NLP. However, when it comes to big data, incorporating sentiment of words into IR models becomes even more important, and as yet no widely accepted standard exists for this task. The contribution of this paper is a framework for quantifying term frequency (TF) variants with sentiments. We propose models derived from the strength of lexical features to improve sentiment-based ranking.
This paper investigates the determinants of financial flexibility of Japanese firms before and after the global financial crisis. In the pre-crisis period, growth opportunity has a positive effect on financial flexibility, but in the post-crisis period, this effect turns to negative and is especially strong for financially constrained firms. These results indicate that in normal time, the Japanese firms pursue financial flexibility for investment demand as argued in previous literature. During difficult time, investment environment deteriorates and investment declines, therefore the firms with less growth opportunity accelerate to accumulate financial flexibility, especially for financially constrained firms as they suffer more than others in such a period. Enhancing the investment environment can improve the efficiency of corporate capital and financial support from banking system may ease the stress of financially constrained firms in post-crisis period.
Finance is not only the lifeblood of an economy, but also the lever to adjust the macro-economy. A modern economy is a market economy and essentially a developed financial economy. Based on the analysis of the problems faced by traditional finance and the overview of smart finance, this study puts forward the application of deep learning combined with reinforcement learning in smart finance to solve the problems existing in financial activities for the first time, and verifies through experiments. The model has better data and information processing ability compared with the traditional financial analysis mode. It provides higher quality decision-making information and bring more benefits. Taking a bond rating report as an example, it usually takes about 2 hours for manual in-depth analysis and carding, while it only takes about 2 minutes to interpret and refine the report by using the deep reinforcement learning model. The model has a certain reference value to solve the problems of traditional finance.
Risk management is an important link in tax administration. From China’s taxation practice, risk identification has become the weakness of tax management. With the complexity of massive data and the secrecy of modern transactions, traditional tax risk identification can no longer adapt to the development of the times. In the past, most risk researches focused on the basic machine learning stage. There are gaps in the application of deep learning in tax risk management. Based on the tax risk management indicators, this paper took the real estate industry as an example. We used convolutional neural network (CNN) to construct a tax risk prediction model. The experiment shows that a tax risk prediction model based on CNN has higher accuracy in tax risk identification and has a stronger ability to process tax data. The model has a certain reference value for tax authorities to reduce tax risk and tax loss.
In this paper we present some computational techniques based on the class of preconditioned Krylov subspace methods that enable us to carry out large-scale, big data simulations of Computational Electromagnetics applications modeled using integral equations. This analysis requires the solution of large linear systems that cannot be afforded by conventional direct methods (based on variants of the Gaussian elimination algorithm) due to their high memory costs. We show that, thanks to the development of efficient Krylov methods and suitable preconditioning techniques, nowadays the solution of realistic electromagnetic problems that involve tens of million (and sometimes even more) unknowns, has become feasible. However, the choice of the best class of methods for the selected computer hardware and the given geometry remains an open problem that requires further analysis.
This article takes the formula of trade complementarity index to calculate the degree of trade complementarity of various industries between China and Nordic countries based on the data of import and export from 2013 to 2018. The results indicate that the trade complementarity between China and Nordic countries has three major characteristics, that is, intensive concentration of specific complementary categories, structurally complementary in bilateral export and import and complementarity based on respective comparative advantages, and clarify the dual complementarities both in inter-industry trade and intra-industry trade. Our findings show that the basis of bilateral cooperation is stable due to strong complementarity each other. In addition, expanding the volume of trade complementary goods and cooperation are not only to promote the development of dual circulation in China, but also to reduce the segmentation costs in the processes of production for both sides.
Tariff relief is a prerequisite for reaching a free trade agreement. This study employs quantitative indicators to identify the specific types of goods with trade complementarity between China and Norway based on the published data by the United Nations, the World Trade Organization and the National Bureaus of statistics both in China and Norway. The empirical results confirm that some types of commodities are provided with trade complementarity between both sides. Nevertheless, these complementary goods are imposed tariffs on each other. The consequences are linked with aggravating resources shortage in China on the one hand, and limiting consumption in Norway on the other. Therefore, pushing up tariff relief is favor of mutual benefit cooperation and making progress in the negotiation of free trade agreement between China and Norway.
This article aims to present with more details, the multicriteria decision aid SAPEVO-M-NC (Simple Aggregation of Preferences Expressed by Ordinal Vectors - Non-Compensatory - Multi Decision Makers). It is a new version of the SAPEVO-M method, of an ordinal, non-compensatory nature and with the possibility of acting by multiple decision makers. As a result, the method provides information on the partial weights, indicating the relative importance of the criteria for each of the decision makers, the relative dominance values and two evaluations on the performance of the alternatives: a partial one, which considers the absolute dominance indices, being used to assess existing dominance relationships; and a global one, which provides the performance rates of the alternatives, making it possible to order them as well as to carry out a sensitivity analysis on the observed performances, reflecting in greater transparency in the decision-making process.
The process of mainframe machines managing and administration requires not only specialized expert knowledge based on many years of experience but also on appropriate tools provided by a machine performance management system, e.g. the Resource Measurement Facility (RMF). The aim of this paper is to show some preliminary results of Z-RAYS system construction that is built basing on machine learning (ML) techniques. It allows automatic detection of anomalies and generation of early warnings about some errors that can appear in the mainframe to support mainframe management process. Presented results are based on extensive simulations that were done basing on the IBM emulator. We focus on determining the degree of the metrics variability, the degree of the data repeatability in metrics, some approaches in metrics anomaly detection and solutions for event correlation detection in metrics.
This paper aims to support the selection decision of a medium-sized warship (between 2,000 and 3,000 tons), to be built in Brazil, presenting the alternatives in a hierarchical manner. Among the various multicriteria decision analysis (MCDA) methods, we used the analytic hierarchy process (AHP) as a basis. Throughout the study, we will propose some adjustments to the AHP in order to make the decision more robust (such as the use of the Gaussian factor and Pearson’s correlation). The criteria were listed and their respective weights were assigned in light of the National Defense Strategy, the Navy’s Strategic Program and interviews carried out with Brazilian navy officers with more than twenty years of career. To list the criteria, we adopted the critical incident technique. The use of the adapted AHP method in choosing the unit to be built can be considered as a transparent way, with a clearly scientific bias, for the Brazilian society to have the perception that the best option was made among the three models of warships presented.
Concerning the expansion of the coronavirus in the world, the search for the development of solutions related to the control and mitigation of the pandemic situation became constant. The paper addresses an analysis of localities for the installation of field hospitals, highly requested as a point of treatment for COVID-19. In this scenario, a framework based on the P-median approach and mathematical programming is proposed, enabling an optimization model as an analysis format for the problematic situation. To support the implementation of the model, a computational tool for data processing was developed, integrating an optimization model to the geographical evaluation, exploring in the analysis numerical and graphical resources. As a validation of the study, a case study in the city of Rio de Janeiro – Brazil is presented, analyzing 162 neighborhoods and determining seven favorable localities for the installation of field hospitals.
Personnel selection is increasingly proving to be an essential factor for the success of organizations. These issues almost universally involve multiple conflicting objectives, uncertainties, costs and benefits in the decision-making process. In this context, due to the characteristics of human resources problems, composed of several subjective evaluation criteria, such problems can be solved by applying Multiple Criteria Decision Aid (MCDA) methods. This paper aims to present a literature review on the main applications of MCDA in the personnel selection area, considering the tactical, operational and strategic spheres. The methodology includes a bibliometric study and a literature review of documents from the Scopus database. We identified the document type, year of publication, affiliation, authors, author’s H-index, the field of knowledge, country and keyword clusters. The literature review allowed us to verify that the Fuzzy Logic is the most applied methodology in personnel selection problems with MCDA, due to its capacity to handle vague, imprecise and subjective data.
3D printing technologies define the essence of Additive Manufacturing and make possible the agile production of customized parts from different materials, with lower unit cost and waste generation. Currently, one of the most widespread 3D printer technologies is the Fused Deposition Modeling (FDM) type, which is the object of this paper. The choice of 3D printing equipment depends on the alignment of the purpose of use and technical knowledge to consider certain requirements. Therefore, this choice can be time-consuming and/or imprecise. In this sense, this work aimed to classify FDM-type 3D printer models by applying the ELECTRE-MOr method, a Multi-criteria Decision Aiding (MCDA) method. As a result, based on a categorization between classes, the FABER 10 alternative was the only one that presented class A performance in all evaluated scenarios, based on criteria defined by the experts consulted in this study.
COVID-19 is a pandemic outbreak for each country worldwide. Each government needs to monitor every citizen and the COVID-19 test becomes an essential evidence for people who are travelling. This gives rise to the necessity of disruptive technologies such as Blockchain. In this paper, we provide an overview of the Hyperledger and Ethereum platforms and present how healthcare organizations can control and monitor digital health test certificates with citizens or other stakeholders. We also present a smart contract structure and implementation for COVID-19 test certificates in both blockchain platforms.
Cybersecurity capacity building has emerged as a notable matter for numerous jurisdictions. Cyber-related threats are posing an ever-greater risk to national security for all countries, irrespective of whether they are developed or in the midst of transitioning. This paper presents the results of two qualitative studies using the Cybersecurity Capacity Maturity Model (CCMM) for nations: (1) Interactive Management (IM) and (2) focus groups to analyse the current state of Spring Land’s cybersecurity capacity. A total of 26 participants from government agencies and five national experts from the Spring Land National Cybersecurity Authority (NCSA) contributed to this study. The results show that Spring Land has many issues such as lack of cybersecurity culture and collaborative road-map across government sectors which results in instability within the country. The assessments feed into the requirement analysis of the National Cybersecurity Capacity Building Framework that can be utilised to organise and test the cybersecurity for nations.
This paper aims to select an algorithm for the Machine Learning (ML) classification task. For the proposed analysis, the Multi-criteria Decision Aid (MCDA) Méthode d’ELimination et de CHoix Includent les relations d’ORdre (MELCHIOR) method was applied. The experiment considered the following criteria as relevant: Accuracy, sensitivity, and processing time of the algorithms. The data used refers to the intention of buying on the Internet and the purpose is to predict whether the customer will finalize a particular purchase. Among various MCDA techniques available, MELCHIOR was chosen to support the decision-making process because this method provides the evaluation of alternatives without the need to elicit the weights of the criteria. As a result, the Gradient Boosting Decision Tree algorithm has been selected as the most suitable for the ML classification task.
The consequences of the pandemic caused by the new coronavirus in the most diverse sectors of the Brazilian economy, are overwhelming, and its effects are still difficult to measure completely. There are several possible scenarios being considered, such as prolonged depression, “U” or “V” recovery. Due to such volatility, risks and uncertainties, the investor, before investing, must carefully analyze the alternatives available in the market. Given the above, this article aims to propose different ways of distributing a financial portfolio, considering five investment funds, which were evaluated in the light of five criteria, by two investors who work in the financial market. Therefore, the SAPEVO-M-NC multicriteria decision aid method was used to evaluate the alternatives, as well as their composition in the investment portfolios. The adoption of the methodology made it possible to carry out the distribution of the portfolio in a clear and consistent way, showing itself as an efficient practical tool for the proposed approach.
This manuscript focuses on the Belt and Road Initiative (BRI) of China, whereby the focus is on the engagement of big data analytics to comprehend logistics exertion. China is the trendsetter for revolutionary practices in trade, logistics, and technology. The recent progress the nation is thriving is on ‘One Belt One Road’ project whereby 65 countries are involved. It aims to connect continents and circulate smooth trade between them. This paper addresses the role of the database to identify the inter-model logistics in BRI. The merits of this project in the perspective of economic growth are measured through a quantitative study with 112 samples. Goal-setting theory is used to construct a conceptual framework for the research. Multivariate analysis is executed with SmartPLS 3.3.3 followed by an in-depth structural equation modeling. Normal distribution of data was given importance as in statistics the real-value of random variables whose distributions are not known, thus Gaussian distribution of data was used. Out of 6 Hypotheses, it is noted that five are significantly positive. Hypothesis testing is concluded based on p-value and t-statistics. The outcome of research suggests that big-data analytics is a major contributor in determining the significant model on logistics in Belt and Road Initiative.
Timing financing theory holds that when listed companies are overvalued, they will increase equity financing. Different from previous studies, this study explores the mechanism of timing financing theory in China from the perspective of financing efficiency. Through the empirical study of M & A cases of Listed Companies in China from 2007 to 2018, it is found that the timing financing behavior that increasing equity financing when overvalued reduces the positive impact of equity financing on M & A performance. Further research found that the mediating effect of the financing efficiency exists, that is, the timing financing behavior does no increase listed companies’ financing efficiency, so it can not help listed companies create value through M & A.
Business leaders around the world are using emerging technologies to capitalize on data, to create business value and to compete effectively in a digitally driven world. Among them the risk assessment and the risk management, based on the assessment is a process which can be made using the available past historical data and applying Data Analytics. Although it is being implemented in different business domains, it is at a nascent stage. It is further new and emerging in the area of Education. This paper describes such a process followed in an educational institution of an engineering college and the use of data for risk management. Based on the processes followed, the performance of the students is seen to be improving in academic performance, placement, higher education and entrepreneurship. This also provides a good process and framework for taking strategic initiatives which will give long term benefits in the areas like research and outreach activities.
Online car-hailing service has had an exponential growth in recent years, and poses a substantial threat to taxi service. Yet how to regulate online car-hailing service has not been adequately studied. Based on the regulation of tradition taxi service, price regulation and entry limitation are used to regulate online car-hailing service in this paper. Moreover, we consider two types of online car-hailing service, i.e., high-end service(e.g. UberBlack) and low-end service (e.g., UberX), according to the perceived service level. Then, the optimal platform’s price is formulated. The result shows that the price regulation are likely to increase the optimal platform price, depending on the service type and the taxi price. When the platform offers low-end service and the taxi price is low, the optimal platform price does not change. In contrast, the entry limitation reduces the optimal platform’s price when it achieves the regulation target.