Congying Guan, Shengfeng Qin and Yang Long
The big challenge in apparel recommendation system research is not the exploration of machine learning technologies in fashion, but to really understand clothes, fashion and…
Abstract
Purpose
The big challenge in apparel recommendation system research is not the exploration of machine learning technologies in fashion, but to really understand clothes, fashion and people, and know what to learn. The purpose of this paper is to explore an advanced apparel style learning and recommendation system that can recognise deep design-associated features of clothes and learn the connotative meanings conveyed by these features relating to style and the body so that it can make recommendations as a skilled human expert.
Design/methodology/approach
This study first proposes a type of new clothes style training data. Second, it designs three intelligent apparel-learning models based on newly proposed training data including ATTRIBUTE, MEANING and the raw image data, and compares the models’ performances in order to identify the best learning model. For deep learning, two models are introduced to train the prediction model, one is a convolutional neural network joint with the baseline classifier support vector machine and the other is with a newly proposed classifier later kernel fusion.
Findings
The results show that the most accurate model (with average prediction rate of 88.1 per cent) is the third model that is designed with two steps, one is to predict apparel ATTRIBUTEs through the apparel images, and the other is to further predict apparel MEANINGs based on predicted ATTRIBUTEs. The results indicate that adding the proposed ATTRIBUTE data that captures the deep features of clothes design does improve the model performances (e.g. from 73.5 per cent, Model B to 86 per cent, Model C), and the new concept of apparel recommendation based on style meanings is technically applicable.
Originality/value
The apparel data and the design of three training models are originally introduced in this study. The proposed methodology can evaluate the pros and cons of different clothes feature extraction approaches through either images or design attributes and balance different machine learning technologies between the latest CNN and traditional SVM.
Details
Keywords
Qing‐Sheng Yang, Cai‐Qin Cui and Xu‐Zhi Lu
The advanced synthetic and natural materials, such as piezoelectric ceramics, electroactive polymers and biological soft tissues, exhibit the multi‐physical or physicochemical…
Abstract
The advanced synthetic and natural materials, such as piezoelectric ceramics, electroactive polymers and biological soft tissues, exhibit the multi‐physical or physicochemical coupling behaviors. The coupling behavior involves the thermal‐mechanical, electric‐mechanical and electrochemicalmechanical interactions. The coupling phenomena can be modeled in the microscopic and macroscopic levels. In the microscale, the material consists of the solid, fluid and ions. The domain FE technique can be used to model the deformation of the solid and the flow of the fluid. In the macroscale, the mixture theory can be applied to description of the coupled response of the continuum under coupled thermal, electrical, chemical and mechanical loadings. A weak form of the governing equations is established by means of variational principle and a multi‐field finite element (MFE) method is developed for numerical modeling of the coupling behavior of advanced materials.
Details
Keywords
Qin Sheng, Fred F. Farshad and Shangyu Duan
In this study, a three‐dimensional (3D) flow model is used to approximate the crystallinity gradients of slowly crystallizing polymers developed in the injection molding process…
Abstract
In this study, a three‐dimensional (3D) flow model is used to approximate the crystallinity gradients of slowly crystallizing polymers developed in the injection molding process. A generalized second order parallel splitting formula is constructed to achieve both the accuracy and efficiency of the computation. Calculated values of flow‐wise (flow‐thickness plane) and width‐wise (width‐thickness plane) crystallinity distributions are obtained and compared with experimental results. The structure‐oriented simulation method developed is not only capable of describing moldability parameters, but is also able to predict the characteristics of ultimate properties of the final products.
Details
Keywords
Qin Sheng, Shekhar Guha and Leonel Gonzalez
The purpose of this paper is to develop highly efficient decomposition finite difference methods for computing solutions of highly oscillatory beam propagation partial…
Abstract
Purpose
The purpose of this paper is to develop highly efficient decomposition finite difference methods for computing solutions of highly oscillatory beam propagation partial differential equations.
Design/methodology/approach
Highly oscillatory optical wave equations, such as the multidimensional paraxial Helmholtz equation, have been used extensively in modelling propagation of the light from lens to the focal region in various engineering applications. Numerical approximations of solutions of such equations contain crucial light information in focal regions even when the f-number is small. However, it has been difficult to acquire highly oscillatory numerical solutions efficiently. This paper proposes two correlated eikonal decomposition strategies for fast computations of the oscillatory solutions. Structures of the numerical methods are designed via an eikonal, or exponential, transformation. The approach converts successfully the oscillatory problems to non-oscillatory subproblems. Therefore, the underlying beam simulation equations can be solved readily with great accuracy and stability.
Findings
It is found that the two correlated eikonal transformation based decomposition methods effectively remove the highly oscillatory features of the wave equations. The coupled non-oscillatory subproblems resulted are easier to solve. Discretization steps in computations can be chosen to be relatively large and this ensures the efficiency of computations. The decomposed finite difference schemes are simple to use in different optical applications.
Practical implications
The computational approach provides a valuable tool to practical applications, such as those in the defence industry.
Originality/value
Although the eikonal transformation has been used in the theory of nonlinear optics, this is the first time it has been utilized for effective engineering computations.
Details
Keywords
Yuanyuan Yin, Shengfeng Qin and Ray Holland
The purpose of this paper is to investigate how to measure collaborative design performance and, in turn, improve the final design output during a design process, with a clear…
Abstract
Purpose
The purpose of this paper is to investigate how to measure collaborative design performance and, in turn, improve the final design output during a design process, with a clear objective to develop a design performance measurement (DPM) matrix to measure design project team members' design collaboration performances.
Design/methodology/approach
The methodology adopted in this research uses critical literature reviews, in‐depth focus group interviews and a questionnaire survey.
Findings
The main finding of this study is a DPM matrix that addresses five DPM indicators: efficiency, effectiveness, collaboration, management skill, and innovation, and 25 detailed DPM criteria. It was found that decision‐making efficiency is the most important DPM criterion for collaborative design efficiency; plus delivering to the brief for effectiveness; clear team goal/objectives for collaboration; decision‐making ability for management skill; and competitive advantage for innovation.
Research limitations/implications
As the present study was focused on exploring DPM during a design process, some key DPM criteria which are not measurable during a design development process were not included in this study. The proposed multi‐feedback approach for DPM matrix implementation needs to be validated in future research.
Practical implications
The DPM matrix can be applied to support a design manager in measuring and improving collaborative design performance during a design process, by reviewing and modifying collaborative design development, identifying the design team strengths and weaknesses, improving team communication, and suggesting suitable responsive actions.
Originality/value
The major contribution of this study is the investigation and development of a DPM matrix to measure collaborative design performance during a design process.
Details
Keywords
Abstract
Purpose
The Internet of Things (IoT) has attracted a lot of attention in both industrial and academic fields for recent years. Artificial intelligence (AI) has developed rapidly in recent years as well. AI naturally combines with the Internet of Things in various ways, enabling big data applications, machine learning algorithms, deep learning, knowledge discovery, neural networks and other technologies. The purpose of this paper is to provide state of the art in AI powered IoT and study smart public services in China.
Design/methodology/approach
This paper reviewed the articles published on AI powered IoT from 2009 to 2018. Case study as a research method has been chosen.
Findings
The AI powered IoT has been found in the areas of smart cities, healthcare, intelligent manufacturing and so on. First, this study summarizes recent research on AI powered IoT systematically; and second, this study identifies key research topics related to the field and real-world applications.
Originality/value
This research is of importance and significance to both industrial and academic fields researchers who need to understand the current and future development of intelligence in IoT. To the best of authors’ knowledge, this is the first study to review the literature on AI powered IoT from 2009 to 2018. This is also the first literature review on AI powered IoT with a case study of smart public service in China.
Details
Keywords
Xianke Sun, Gaoliang Wang, Liuyang Xu and Honglei Yuan
In data grids, replication has been regarded as a crucial optimization strategy. Computing tasks are performed on IoT gateways at the cloud edges to obtain a prompt response. So…
Abstract
Purpose
In data grids, replication has been regarded as a crucial optimization strategy. Computing tasks are performed on IoT gateways at the cloud edges to obtain a prompt response. So, investigating the data replication mechanisms in the IoT is necessary. Henceforth, a systematic survey of data replication strategies in IoT techniques is presented in this paper, and some suggestions are offered for the upcoming works. In two key classifications, various parameters dependent on the analysis of the prevalent approaches are considered. The pros and cons associated with chosen strategies have been explored, and the essential problems of them have been presented to boost the future of more effective data replication strategies. We have also discovered gaps in papers and provided solutions for them.
Design/methodology/approach
Progress in Information Technology (IT) growth has brought the Internet of Things (IoT) into life to take a vital role in our everyday lifestyles. Big IoT-generated data brings tremendous data processing challenges. One of the most challenging problems is data replication to improve fault-tolerance, reliability, and accessibility. In this way, if the primary data source fails, a replica can be swapped in immediately. There is a significant influence on the IoT created by data replication techniques, but no extensive and systematic research exists in this area. There is still no systematic and full way to address the relevant methods and evaluate them. Hence, in the present investigation, a literature review is indicated on the IoT-based data replication from papers published until 2021. Based on the given guidelines, chosen papers are reviewed. After establishing exclusion and inclusion criteria, an independent systematic search in Google Scholar, ACM, Scopus, Eric, Science Direct, Springer link, Emerald, Global ProQuest, and IEEE for relevant studies has been performed, and 21(6 paper analyzed in section 1 and 15 paper analyzed in section 3) papers have been analyzed.
Findings
The results showed that data replication mechanisms in the IoT algorithms outperform other algorithms regarding impressive network utilization, job implementation time, hit ratio, total replication number, and the portion of utilized storage in percentage. Although a few ideas have been suggested that fix different facets of IoT data management, we predict that there is still space for development and more study. Thus, in order to design innovative and more effective methods for future IoT-based structures, we explored open research directions in the domain of efficient data processing.
Research limitations/implications
The present investigation encountered some drawbacks. First of all, only certain papers published in English were included. It is evident that some papers exist on data replication processes in the IoT written in other languages, but they were not included in our research. Next, the current report has only analyzed the mined based on data replication processes and IoT keyword discovery. The methods for data replication in the IoT would not be printed with keywords specified. In this review, the papers presented in national conferences and journals are neglected. In order to achieve the highest ability, this analysis contains papers from major global academic journals.
Practical implications
To appreciate the significance and accuracy of the data often produced by different entities, the article illustrates that data provenance is essential. The results contribute to providing strong suggestions for future IoT studies. To be able to view the data, administrators have to modify novel abilities. The current analysis will deal with the speed of publications and suggest the findings of research and experience as a future path for IoT data replication decision-makers.
Social implications
In general, the rise in the knowledge degree of scientists, academics, and managers will enhance administrators' positive and consciously behavioral actions in handling IoT environments. We anticipate that the consequences of the present report could lead investigators to produce more efficient data replication methods in IoT regarding the data type and data volume.
Originality/value
This report provides a detailed literature review on data replication strategies relying on IoT. The lack of such papers increases the importance of this paper. Utilizing the responses to the study queries, data replication's primary purpose, current problems, study concepts, and processes in IoT are summarized exclusively. This approach will allow investigators to establish a more reliable IoT technique for data replication in the future. To the best of our understanding, our research is the first to provide a thorough overview and evaluation of the current solutions by categorizing them into static/dynamic replication and distributed replication subcategories. By outlining possible future study paths, we conclude the article.
Details
Keywords
The purpose of this paper is to present a detailed and critical look at the evolution of online survey research since Evans and Mathur’s (2005) article on the value of online…
Abstract
Purpose
The purpose of this paper is to present a detailed and critical look at the evolution of online survey research since Evans and Mathur’s (2005) article on the value of online surveys. At that time, online survey research was in its early stages. Also covered are the present and future states of online research. Many conclusions and recommendations are presented.
Design/methodology/approach
The look back focuses on online surveys, strengths and weaknesses of online surveys, the literature on several aspects of online surveys and online survey best practices. The look ahead focuses on emerging survey technologies and methodologies, and new non-survey technologies and methodologies. Conclusions and recommendations are provided.
Findings
Online survey research is used more frequently and better accepted by researchers than in 2005. Yet, survey techniques are still regularly transformed by new technologies. Non-survey digital research is also more prominent than in 2005 and can better track actual behavior than surveys can. Hybrid surveys will be widespread in the future.
Practical implications
The paper aims to provide insights for researchers with different levels of online survey experience. And both academics and practitioners should gain insights.
Social implications
Adhering to a strong ethics code is vital to gain respondents’ trust and to produce valid results.
Originality/value
Conclusions and recommendations are offered in these specific areas: defining concepts, understanding the future role of surveys, developing and implementing surveys and a survey code of ethics. The literature review cites more than 200 sources.
Details
Keywords
Mohamad Abu Ghazaleh and Abdelrahim M. Zabadi
Internet of things (IoT) and big data (BD) could change how the societies function. This paper explores the role of IoT and BD and their impact on customer relationship management…
Abstract
Purpose
Internet of things (IoT) and big data (BD) could change how the societies function. This paper explores the role of IoT and BD and their impact on customer relationship management (CRM) investments in modern customer service. The purpose of this paper is to develop an analytic hierarchy planning framework to establish criteria weights and to develop a general self-assessment model for determining the most important factors influencing the IoT and BD investment in CRM. The authors found that most studies have focused on conceptualizing the impact of IoT without BD and with limited empirical studies and analytical models. This paper sheds further light on the topic by presenting both IoT and BD aspects of future CRM.
Design/methodology/approach
The analytic hierarchy process (AHP) methodology is used to weight and prioritize the factors influencing the IoT and BD investment in modern CRM in the service industry. The AHP framework resulted in a ranking of 21 sustainability sub-factors based on evaluations by experienced information technology and customer service professionals.
Findings
The paper provides significant insight on the new frontier of CRM, focusing on the use of IoT and BD and the respective solutions to address them were identified. This study primarily contributes in providing the process of effectively managing and implementing IoT and BD in big businesses by identifying the connecting link between firms and customers.
Practical implications
The understanding of new frontier of CRM connective via IoT and BD can solve the dilemmas and challenges linked to the practice of implement IoT and BD in the information systems field. The study provides valuable information and critical analysis of IoT and BD with regard to the integration of CRM. Finally, this study further provides directions for future researchers.
Originality/value
IoT and BD are a growing phenomenon, which business decision-makers and information professionals need to consider seriously to properly ascertain the modern CRM dimensions in the digital economies. They also should embrace the proper CRM innovation, which is powered by IoT and BD, and discover how IoT and BD can bring the next level of maturity to CRM “CRM of everything.”
Details
Keywords
Efthimia Mavridou, Konstantinos M. Giannoutakis, Dionysios Kehagias, Dimitrios Tzovaras and George Hassapis
Semantic categorization of Web services comprises a fundamental requirement for enabling more efficient and accurate search and discovery of services in the semantic Web era…
Abstract
Purpose
Semantic categorization of Web services comprises a fundamental requirement for enabling more efficient and accurate search and discovery of services in the semantic Web era. However, to efficiently deal with the growing presence of Web services, more automated mechanisms are required. This paper aims to introduce an automatic Web service categorization mechanism, by exploiting various techniques that aim to increase the overall prediction accuracy.
Design/methodology/approach
The paper proposes the use of Error Correcting Output Codes on top of a Logistic Model Trees-based classifier, in conjunction with a data pre-processing technique that reduces the original feature-space dimension without affecting data integrity. The proposed technique is generalized so as to adhere to all Web services with a description file. A semantic matchmaking scheme is also proposed for enabling the semantic annotation of the input and output parameters of each operation.
Findings
The proposed Web service categorization framework was tested with the OWLS-TC v4.0, as well as a synthetic data set with a systematic evaluation procedure that enables comparison with well-known approaches. After conducting exhaustive evaluation experiments, categorization efficiency in terms of accuracy, precision, recall and F-measure was measured. The presented Web service categorization framework outperformed the other benchmark techniques, which comprise different variations of it and also third-party implementations.
Originality/value
The proposed three-level categorization approach is a significant contribution to the Web service community, as it allows the automatic semantic categorization of all functional elements of Web services that are equipped with a service description file.