Abstract Computational techniques for network clustering identification are critical to several application domains. Recently, Modularity Maximization and Modularity Density Maximization have become two of the main techniques that provide computational methods to identify network clusterings. Therefore, understanding their differences and common characteristics is fundamental to decide which one is best suited for a given application. Several heuristics and exact methods have been developed for both Modularity Maximization and Modularity Density Maximization problems. Unfortunately, no structured methodological comparison between the two techniques has been proposed yet. This paper reports a ground truth contest between both optimization problems. We do so aiming to compare their exact solutions and the results of heuristics inspired in these problems. In our analysis, we use branch-and-price exact methods which apply the best-known column generation procedures. The heuristic methods obtain the highest objective function scores and find solutions for networks with hundreds of thousands of nodes. Our experiments suggest that Modularity Density Maximization yields the best results over the tested networks. The experiments also show the behavior and importance of the quantitative factor of the Modularity Density Maximization objective function.

With the rapid growth of social networks, mining customer opinions based on online reviews is crucial to understand consumer needs. Due to the richness of language expressions, customer opinions are often expressed implicitly. However, previous studies usually focus on mining explicit opinions to understand consumer needs. In this paper, we propose a novel implicit opinion analysis model to perform implicit opinion analysis of Chinese customer reviews at both the feature and review levels. First, we extract an implicit-opinionated review/clause dataset from raw review dataset and introduce the concept of the feature-based implicit opinion pattern (FBIOP). Secondly, we develop a clustering algorithm to construct product feature categories. Based on the constructed feature categories, FBIOPs can be mined from the extracted implicit-opinionated clause dataset. Thirdly, the sentiment intensity and polarity of each FBIOP are calculated by using the Chi squared test and pointwise mutual information. Fourthly, according to the resulting FBIOP polarities, the polarities of implicit opinions can be determined at both the feature and review levels. Car forum reviews written in Chinese are collected and labeled as the experimental dataset. The results show that the proposed model outperforms the traditional support vector machine model and the cutting-edge convolutional neural network model.

Abstract The massive growth of data in recent years has led challenges in data mining and machine learning tasks. One of the major challenges is the selection of relevant features from the original set of available features that maximally improves the learning performance over that of the original feature set. This issue attracts researchers’ attention resulting in a variety of successful feature selection approaches in the literature. Although there exist several surveys on unsupervised learning (e.g., clustering), lots of works concerning unsupervised feature selection are missing in these surveys (e.g., evolutionary computation based feature selection for clustering) for identifying the strengths and weakness of those approaches. In this paper, we introduce a comprehensive survey on feature selection approaches for clustering by reflecting the advantages/disadvantages of current approaches from different perspectives and identifying promising trends for future research.

The objective of this work is to present a novel multi-attribute decision making (MADM) method under interval-valued intuitionistic fuzzy (IVIF) set environment by integrating a Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method. Set pair analysis (SPA) theory is the modern uncertainty theory which is composed by the three components, namely “identity”, “discrepancy” and “contrary” degrees of the connection number (CN) and overlap with the various existing theories for handling the uncertainties in the data. Thus, motivated by this, in the present work, an attempt is made to enrich the theory of information measure by presented some exponential based distance measures using CNs of the IVIF sets. The supremacy of the proposed measure is also discussed. Afterward, a TOPSIS method based on the proposed distance measures is developed to solve MADM problem under IVIF environment where each of the element is characteristics by IVIF numbers. The utility, as well as supremacy of the approach, is confirmed through a real-life numerical example and validate it by comparing their results with the several existing approaches results.

Bursts in human and natural activities are highly clustered in time or space, suggesting that these activities are influenced by previous events within the social or natural system. Such bursty behavior in the real world conveys substantial information of underlying diffusion processes, which have been studied based on point process approaches in diverse scientific communities from online social media to criminology and epidemiology. However, universal components of real-world diffusion dynamics that cut across disciplines remain unexplored with a common overarching perspective. In this review, we introduce a wide range of diffusion processes from diverse research fields, define a taxonomy of common major factors in diffusion dynamics, interpret their diffusion models from the theoretical perspectives of point processes, and compare them with respect to universal effects on diffusion. These all can provide new insights on spatial and temporal bursty events capturing underlying diffusion dynamics. We expect that the comprehensive aspects of diffusion dynamics in the real world can motivate transdisciplinary research and provide contextual components of a fundamental framework for more generalizable diffusion models.

Abstract Collaborative filtering recommender systems (CFRSs) have already been proved effective to cope with the information overload problem since they merged in the past two decades. However, CFRSs are highly vulnerable to shilling or profile injection attacks since their openness. Ratings injected by malicious users seriously affect the authenticity of the recommendations as well as users’ trustiness in the recommendation systems. In the past two decades, various studies have been conducted to scrutinize different profile injection attack strategies, shilling attack detection schemes, robust recommendation algorithms, and to evaluate them with respect to accuracy and robustness. Due to their popularity and importance, we survey about shilling attacks in CFRSs. We first briefly discuss the related survey papers about shilling attacks and analyze their deficiencies to illustrate the necessity of this paper. Next we give an overall picture about various shilling attack types and their deployment modes. Then we explain profile injection attack strategies, shilling attack detection schemes and robust recommendation algorithms proposed so far in detail. Moreover, we briefly explain evaluation metrics of the proposed schemes. Last, we discuss some research directions to improve shilling attack detection rates robustness of collaborative recommendation, and conclude this paper.

In this paper, we introduce a methodology for improving the targeting of marketing campaigns using bridge prediction in communities based on the information of multilayer online social networks. The campaign strategy involves the identification of nodes with high brand loyalty and top-ranking nodes in terms of participation in bridges that will be involved in the evolution of the graph. Our approach is based on an efficient classification model combining topological characteristics of crawled social graphs with sentiment and linguistic traits of user-nodes, popularity in social media as well as meta path-based features of multilayer networks. To validate our approach we present a set of experimental results using a well-defined dataset from Twitter and Foursquare. Our methodology is useful to recommendation systems as well as to marketers who are interested to use social influence and run effective marketing campaigns.

Abstract Helpfulness of online reviews is a multi-faceted concept. The reviews are usually ranked on the basis of perceived helpful votes and aid in making purchase decisions for online customers. This study extends the prior work done for review helpfulness by considering not only the influential characteristics of reviews but also incorporates influential indicators of reviewer and product category. Influential factor based new features (product, reviewer and review) are proposed to predict the helpfulness of online reviews by using five ML methods. The experimental analysis on a real-life review dataset shows that the hybrid set of proposed features deliver the best predictive performance. In addition, the reviewer and the review category features introduced in this research exhibit better predictive performance as a standalone model. Findings show that reviews which have large number of comments, large values of sentiment and polarity scores receive more helpful votes. The reviewer activity length and recency are statistically significant predictors for helpfulness prediction. In addition, number of question answered, ratio of positive reviews and average rating per review are also significant variables of product type. The findings of this study highlight the number of implications for research and provide new insights to retailers for efficient ranking and organization of consumer reviews for online users.

Abstract Identify the application of computational methods and algorithms reported in the literature based on four main categories including data mining, clinical decision support systems, geographical information systems, and digital dashboards and to summarize them in a qualitative scoping review. A scoping review was presented following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. MEDLINE, Emerald, Scopus, and Google Scholar databases were searched in July 2016 using uniform keywords for documents that discuss data mining and knowledge discovery, dashboards, geographical information systems, and electronic surveillance of antimicrobial resistance in surveillance systems. Our study mainly focused on knowledge discovery and visualization algorithms, methods, and techniques used in antimicrobial resistance surveillance systems. Thirteen of the reviewed articles applied algorithms to the data mining process. A comparative table of data elements in the reviewed studies was extracted. The characteristics of antimicrobial dashboards were discussed. Heat maps were the most popular method used to visualize the intensity of resistance. Comparative tables are provided in each section of this paper. Data mining, Decision Support Systems, Geographic Information Systems, and dashboards can be integrated for data analysis and to better solve decision support problems. Bio-surveillance systems should be designed and analyzed based on four categories: data mining, dashboards, geography information system, and decision support modules. Furthermore, some questionnaires and checklists were developed and validated to capture related Business Intelligence and analytical requirements. Future studies should focus on developing fast, flexible, and accurate computational bio-surveillance systems by appropriate selecting and applying the considered methods and algorithms.

Abstract In this paper we present a broad overview of the last 40 years of research on cognitive architectures. To date, the number of existing architectures has reached several hundred, but most of the existing surveys do not reflect this growth and instead focus on a handful of well-established architectures. In this survey we aim to provide a more inclusive and high-level overview of the research on cognitive architectures. Our final set of 84 architectures includes 49 that are still actively developed, and borrow from a diverse set of disciplines, spanning areas from psychoanalysis to neuroscience. To keep the length of this paper within reasonable limits we discuss only the core cognitive abilities, such as perception, attention mechanisms, action selection, memory, learning, reasoning and metareasoning. In order to assess the breadth of practical applications of cognitive architectures we present information on over 900 practical projects implemented using the cognitive architectures in our list. We use various visualization techniques to highlight the overall trends in the development of the field. In addition to summarizing the current state-of-the-art in the cognitive architecture research, this survey describes a variety of methods and ideas that have been tried and their relative success in modeling human cognitive abilities, as well as which aspects of cognitive behavior need more research with respect to their mechanistic counterparts and thus can further inform how cognitive science might progress.

In the case of vector data, Gretton et al. (Algorithmic learning theory. Springer, Berlin, pp 63–77, 2005) defined Hilbert–Schmidt independence criterion, and next Cortes et al. (J Mach Learn Res 13:795–828, 2012) introduced concept of the centered kernel target alignment (KTA). In this paper we generalize these measures of dependence to the case of multivariate functional data. In addition, based on these measures between two kernel matrices (we use the Gaussian kernel), we constructed independence test and nonlinear canonical variables for multivariate functional data. We show that it is enough to work only on the coefficients of a series expansion of the underlying processes. In order to provide a comprehensive comparison, we conducted a set of experiments, testing effectiveness on two real examples and artificial data. Our experiments show that using functional variants of the proposed measures, we obtain much better results in recognizing nonlinear dependence.

In recent times, technological advancement boosts the desire of utilizing the autonomous Unmanned Aerial Vehicle (UAV) in both civil and military sectors. Among various UAVs, the ability of rotary wing UAVs (RUAVs) in vertical take-off and landing, to hover and perform quick maneuvering attract researchers to develop models fully autonomous control framework. The majority of first principle techniques in modeling and controlling RUAV face challenges in incorporating and handling various uncertainties. Recently various fuzzy and neuro-fuzzy based intelligent systems are utilized to enhance the RUAV’s modeling and control performance. However, the majority of these fuzzy systems are based on batch learning methods, have static structure, and cannot adapt to rapidly changing environments. The implication of Evolving Intelligent System based model-free data-driven techniques can be a smart option since they can adapt their structure and parameters to cope with sudden changes in the behavior of RUAVs real-time flight. They work in a single pass learning fashion which is suitable for online real-time deployment. In this paper, state of the art of various fuzzy systems from the basic fuzzy system to evolving fuzzy system, their application in a RUAV namely quadcopter with existing limitations, and possible opportunities are analyzed. Besides, a variety of first principle techniques to control the quadcopter, their impediments, and conceivable solution with recently employed evolving fuzzy controllers are reviewed.

Abstract Search engine optimization (SEO) has been considered one of the most important techniques in internet marketing. This study establishes a decision model of search engine ranking for administrators to improve the performances of websites that satisfy users’ needs. To probe into the interrelationship and influential weights among criteria of SEO and evaluate the gaps of performance to achieve the aspiration level in real world, this research utilizes hybrid modified multiple criteria decision-making models, including decision-making trial and evaluation laboratory (DEMATEL), DEMATEL-based analytic network process (called DANP), and VlseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR). The empirical findings discover that the criteria of SEO possessed a self-effect relationship based on DEMATEL technique. According to the influential network relation map (INRM), external website optimization is the top priority dimension that needs to be improved when implementing SEO. Among the six criteria for evaluation, meta tags is the most significant criterion influencing search engine ranking, followed by keywords and website design. The evaluation of search engine ranking reveals that the website with lowest gap would be the optimal example for administrators of websites to make high ranking website during the time that this study is executed.

Abstract Innovative technology on intelligent processes for smart home applications that utilize Internet of Things (IoT) is mainly limited and dispersed. The available trends and gaps were investigated in this study to provide valued visions for technical environments and researchers. Thus, a survey was conducted to create a coherent taxonomy on the research landscape. An extensive search was conducted for articles on (a) smart homes, (b) IoT and (c) applications. Three databases, namely, IEEE Explore, ScienceDirect and Web of Science, were used in the article search. These databases comprised comprehensive literature that concentrate on IoT-based smart home applications. Subsequently, filtering process was achieved on the basis of intelligent processes. The final classification scheme outcome of the dataset contained 40 articles that were classified into four classes. The first class includes the knowledge engineering process that examines data representation to identify the means of accomplishing a task for IoT applications and their utilisation in smart homes. The second class includes papers on the detection process that uses artificial intelligence (AI) techniques to capture the possible changes in IoT-based smart home applications. The third class comprises the analytical process that refers to the use of AI techniques to understand the underlying problems in smart homes by inferring new knowledge and suggesting appropriate solutions for the problem. The fourth class comprises the control process that describes the process of measuring and instructing the performance of IoT-based smart home applications against the specifications with the involvement of intelligent techniques. The basic features of this evolving approach were then identified in the aspects of motivation of intelligent process utilisation for IoT-based smart home applications and open-issue restriction utilisation. The recommendations for the approval and utilisation of intelligent process for IoT-based smart home applications were also determined from the literature.

Abstract Social Book Search is a new area of social search. In the modern world everything is going to be amenable due to the web and social media. The web and social media give us access to a wealth of information, not only different in quantity but also in character. Traditional descriptions from professionals are now supplemented with user generated content. Although books have been the predominant source of information for centuries, the way we acquire, share, and publish information has changed and has been changing in fundamental ways due to the web. In the modern era, in order to purchase a book, the users are not only depend on the title, author name, publisher etc of the book available as controlled metadata but also on other aspects which include the reviews, editorial reviews etc available in different social media. Our primary focus in this survey is to describe the features of different social cataloging book sites as well as their recommendation based on books. How the online searching of books is useful to the user and up to what extent, and what are their aim are some of the issues we shall deal with. We will also discuss evolution of those techniques for Social Book Search presented over years at the Initiative for the Evaluation of XML Retrieval (INEX) and Conference and Labs of the Evaluation Forum (CLEF). To what extent the features and functionality of a social sharing platform influence the user behavior, is also discussed in the paper. This survey provides an overview of research done in the area of Social Book Search from perspective of Information Retrieval.

Intelligent scheduling covers various tools and techniques for successfully and efficiently solving the scheduling problems. In this paper, we provide a survey of intelligent scheduling systems by categorizing them into five major techniques containing fuzzy logic, expert systems, machine learning, stochastic local search optimization algorithms and constraint programming. We also review the application case studies of these techniques.

Abstract Radiologists interprets X-ray samples by visually inspecting them to diagnose the presence of fractures in various bones. Interpretation of radiographs is a time-consuming and intense process involving manual examination of fractures. In addition, clinician’s shortage in medically under-resourced areas, unavailability of expert radiologists in busy clinical settings or fatigue caused due to demanding workloads could lead to false detection rate and poor recovery of the fractures. A comprehensive study is imparted here covering fracture diagnosis with the aim to assist investigators in developing models that automatically detects fracture in human bones. The paper is presented in five folds. Firstly, we discuss data preparation stage. Second, we present various image-processing techniques used for fracture detection. Third, we analyze conventional and deep learning based techniques for diagnosing bone fractures. Fourth, we make comparative analysis of existing techniques. Fifth, we discuss different issues and challenges faced by researches while dealing with fracture detection.

A logistics center is the hub of a specific area, within various logistics-related activities (distribution, storage, transportation, consolidation, handling, customs clearance, imports, exports, transit processes, infrastructural services, insurance, banking, etc.) that are performed on a commercial basis. Determining the location of the logistics center is an important decision regarding cost and benefit analysis. A three-stage methodology has been applied for presenting a framework for logistics center location selection in the context of Kayseri’s logistics development plan. The first stage includes the determination of criteria through literature review and interviews with experts. The second stage includes the weighting of determined criteria using linear BWM (best–worst method). The third stage includes the ranking of locations using the evaluation based on distance from average solution (EDAS) method with different distance measures. Our proposed methodology BWM–EDAS and also EDAS with different distance measures, which are applied for the first time in the literature, provides helpful findings to rank the logistics center locations. Lastly, sensitivity analysis is conducted to validate the robustness.

Covering based intuitionistic fuzzy (IF) rough set is a generalization of granular computing and covering based rough sets. By combining covering based rough sets, IF sets and fuzzy rough sets, we introduce three classes of coverings based IF rough set models via IF\(\beta \)-neighborhoods and IF complementary \(\beta \)-neighborhood (IFC\(\beta \)-neighborhood). The corresponding axiomatic systems are investigated, respectively. In particular, the rough and precision degrees of covering based IF rough set models are discussed. The relationships among these types of coverings based IF rough set models and covering based IF rough set models proposed by Huang et al. (Knowl Based Syst 107:155–178, 2016). Based on the theoretical analysis for coverings based IF rough set models, we put forward intuitionistic fuzzy TOPSIS (IF-TOPSIS) methodology to multi-attribute decision-making (MADM) problem with the evaluation of IF information problem. An effective example is to illustrate the proposed methodology. Finally, we deal with MADM problem with the evaluation of fuzzy information based on CFRS models. By comparative analysis, we find that it is more effective to deal with MADM problem with the evaluation of IF information based on CIFRS models than the one with the evaluation of fuzzy information based on CFRS models.

Abstract Neutrosophic set, initiated by Smarandache, is a novel tool to deal with vagueness considering the truth-membership T, indeterminacy-membership I and falsity-membership F satisfying the condition \(0\le T+I+F\le 3\). It can be used to characterize the uncertain information more sufficiently and accurately than intuitionistic fuzzy set. Neutrosophic set has attracted great attention of many scholars that have been extended to new types and these extensions have been used in many areas such as aggregation operators, decision making, image processing, information measures, graph and algebraic structures. Because of such a growth, we present an overview on neutrosophic set with the aim of offering a clear perspective on the different concepts, tools and trends related to their extensions. A total of 137 neutrosophic set publication records from Web of Science are analyzed. Many interesting results with regard to the annual trends, the top players in terms of country level as well as institutional level, the publishing journals, the highly cited papers, and the research landscape are yielded and explained in-depth. The results indicate that some developing economics (such as China, India, Turkey) are quite active in neutrosophic set research. Moreover, the co-authorship analysis of the country and institution, the co-citation analysis of the journal, reference and author, and the co-occurrence analysis of the keywords are presented by VOSviewer software.

Metalearning attracted considerable interest in the machine learning community in the last years. Yet, some disagreement remains on what does or what does not constitute a metalearning problem and in which contexts the term is used in. This survey aims at giving an all-encompassing overview of the research directions pursued under the umbrella of metalearning, reconciling different definitions given in scientific literature, listing the choices involved when designing a metalearning system and identifying some of the future research challenges in this domain.