Search

Detecting product adoption intentions on social media could yield significant value in a wide range of applications, such as personalized recommendations and targeted marketing. In the literature, no study has explored the detection of product adoption intentions on social media, and only a few relevant studies have focused on purchase intention detection for products in one or several categories. Focusing on a product category rather than a specific product is too coarse-grained for precise advertising. Additionally, existing studies primarily focus on using one type of text representation in target social media posts, ignoring the major yet unexplored potential of fusing different text representations. In this paper, we first formulate the problem of product adoption intention mining and demonstrate the necessity of studying this problem and its practical value. To detect a product adoption intention for an individual product, we propose a novel and general multiview deep learning model that simultaneously taps into the capability of multiview learning in leveraging different representations and deep learning in learning latent data representations using a flexible nonlinear transformation. Specifically, the proposed model leverages three different text representations from a multiview perspective and takes advantage of local and long-term word relations by integrating convolutional neural network (CNN) and long short-term memory (LSTM) modules. Extensive experiments on three Twitter datasets demonstrate the effectiveness of the proposed multiview deep learning model compared with the existing benchmark methods. This study also significantly contributes research insights to the literature about intention mining and provides business value to relevant stakeholders such as product providers.

PurposeThis research aims to uncover coronavirus disease 2019’s (COVID-19's) impact on shipping and logistics using Internet articles as the source.Design/methodology/approachThis research applies web mining to collect information on COVID-19's impact on shipping and logistics from Internet articles. The information extracted is then analyzed through machine learning algorithms for useful insights.FindingsThe research results indicate that the recovery of the global supply chain in China could potentially drive the global supply chain to return to normalcy. In addition, researchers and policymakers should prioritize two aspects: (1) Ease of cross-border trade and logistics. Digitization of the supply chain and applying breakthrough technologies like blockchain and IoT are needed more than ever before. (2) Supply chain resilience. The high dependency of the global supply chain on China sounds like an alarm of supply chain resilience. It calls for a framework to increase global supply chain resilience that enables quick recovery from disruptions in the long term.Originality/valueDiffering from other studies taking the natural language processing (NLP) approach, this research uses Internet articles as the data source. The findings reveal significant components of COVID-19's impact on shipping and logistics, highlighting crucial agendas for scholars to research.

The popularity of decision tree forest is due to its superior performance and accuracy. The accuracy of the decision tree forest algorithm depends upon the base learner and its diversification. As ...

The paper provides a brief review of current trends in studying ultra-right radicalization risks both in Russia and globally. Since the scientific interpretations in studying the notion of radicalization are differentiated, the authors prefer the following one: the ultra-rightists represent communities and movements that accept the idea that violence is necessary to achieve any goal (political, ideological, economic, social or personal). The ultra-rightists justify and promote this idea, expressing their willingness to act violently. They also make a moral commitment to defend those who promote the idea. The authors present the results of the work of the TPU’s cross-subject project team to create a prototype and a method for automated detection of ultra-rightists’ threats in social media. The paper describes the main challenges the researches face when applying smart social media content analysis as a tool for automating social science research.

İnternet üzerinde muazzam büyüklükte veri barındırmaktadır. Bu veriler istenirse web madenciliği/kazıma olarak adlandırılan yöntem ile elde edilebilmekte ve bilgi çıkarımında kullanılabilmektedir. Araştırma, web madenciliği kullanarak Türk sineması hakkında veri elde etmek ve bu verileri metin madenciliği ile analiz etmenin bir sonucu olarak Türk sineması hakkında çeşitli bulgular elde etmek amacıyla gerçekleştirilmiştir. Toplam 8596 TV ve sinema filmi üzerinde yapılan metin madenciliğinde, en yüksek izleyici puanının, Hababam Sınıfı (1975, p. = 9,3), Tosun Paşa (1976, p. = 9,0), 25 Litre (2019, p. = 8,9) filmlerine ait olduğu görülmektedir. En yüksek puanlı 20 filmden 13 tanesinin Türk Sineması’nda Yeşilçam olarak adlandırılan dönemde yapıldığı görülmektedir. Filmlerden 13 tanesi Komedi türünde yapılmıştır. Listede yönetmen olarak Ertem Eğilmez’in 8, Kartal Tibet’in 2, Atıf Yılmaz’ın 2 filmi yer almaktadır. 1950’den sonra Türk Sinemasında en çok film çekilen yıl 1975 yılı olurken (275 film), en az film çekilen yıl 2002 olmuştur (20 film). Web madenciliği yöntemi internet üzerindeki verilerin robot yazılımlar veya programlama dilleri ile elde edilmesi konusunda birçok alanda kullanılabilecek alternatif bir veri elde etme yöntemi sunmaktadır. Bu yöntem, Türk Sineması’na ait çeşitli bilgileri elde etmede örnek olarak kullanılmıştır. İnternet büyük veri kaynakları ile bilimsel araştırmalarda kullanılabilecek zengin içeriğe sahiptir., It contains an enormous amount of data on the Internet. If desired, these data can be obtained by a method called web mining/scraping and can be used in information extraction. This research was carried out to obtain data about Turkish cinema by using web mining and to obtain various findings about Turkish cinema as a result of analyzing this data with text mining.In the text mining conducted on a total of 8596 TV and cinema films, it is seen that highest audience score belongs to the films Hababam Sınıfı (1975, p. = 9,3), Tosun Paşa (1976, p. = 9,0), 25 Liters (2019, p. = 8,9), respectively. It is seen that 13 of the 20 highest-rated films were made in the period called Yeşilçam in Turkish Cinema. 13 of these films were made in the Comedy genre. The list includes 8 films by Ertem Eğilmez, 2 films by Kartal Tibet and 2 films by Atıf Yılmaz as directors. After 1950, the year in which the most films were made in Turkish cinema was 1975 (275 films), while the year in which the fewest films were made was 2002 (20 films). Web mining method offers an alternative data acquisition method that can be used in many areas to obtain data on the internet with robot software or programming languages. This method has been used as an example to obtain various information about Turkish Cinema. The Internet has rich content that can be used in scientific research with big data sources.

Governmental establishments are maintaining historical data for job applicants for future analysis of predication, improvement of benefits, profits, and development of organizations and institutions. In e-government, a decision can be made about job seekers after mining in their information that will lead to a beneficial insight. This paper proposes the development and implementation of an applicant's appropriate job prediction system to suit his or her skills using web content classification algorithms (Logit Boost, j48, PART, Hoeffding Tree, Naive Bayes). Furthermore, the results of the classification algorithms are compared based on data sets called "job classification data" sets. Experimental results indicated that the algorithm j48 had the highest precision (94.80%) compared to other algorithms for the aforementioned dataset.

Aging in place has garnered a lot of interest in the past decade among researchers and politicians. It presents itself as a humane and cost-efficient solution to the worsening problem related to the financing and staffing of our fragile healthcare ecosystem. In that regard, smart environments could help monitor the daily activities of the person and provide information regarding the health status and any potential problem necessitating immediate assistance. To do so, many teams, including ours, have been working on human activity recognition from distributed sensors. Nevertheless, it is still very challenging due to the difficulty of amassing enough data to learn activity models that generalizes well across different residents and different smart environments. Moreover, whenever one wants to add a new activity to the set of recognizable activities, it requires to gather additional data with label. The whole process is prohibitively costly and time consuming. Therefore, in this paper, our team proposed to explore web mining in order to build activity model in an unsupervised fashion. More specifically, using the results of two popular image search engines with automated queries and with the help of object detection/classification models, we learned the set of key objects involved in the realization of activities of daily life. A total of 108 configurations were tested. The experiments showed that the key objects related to activities can be easily extracted with a good accuracy.

U.S. companies have a growing demand for innovative employees with creative thinking and innovative problem-solving skills. This study explored how the Association to Advance Collegiate Schools of ...

Due to the daily expansion of the web, the amount of information has increased significantly. Thus, the need for retrieving relevant information has also increased. In order to explore the internet, users depend on various search engines. Search engines face a significant challenge in returning the most relevant results for a user's query. The search engine's performance is determined by the algorithm used to rank web pages, which prioritizes the pages with the most relevancy to appear at the top of the result page. In this paper, various web page ranking algorithms such as Page Rank, Time Rank, EigenRumor, Distance Rank, SimRank, etc. are analyzed and compared based on some parameters, including the mining technique to which the algorithm belongs (for instance, Web Content Mining, Web Structure Mining, and Web Usage Mining), the methodology used for ranking web pages, time complexity (amount of time to run an algorithm), input parameters (parameters utilized in the ranking process such as InLink, OutLink, Tag name, Keyword, etc.), and the result relevancy to the user query.

[EN] The main content of a webpage is often surrounded by other boilerplate elements related to the template, such as menus, advertisements, copyright notices, and comments. For crawlers and indexers, isolating the main content from the template and other noisy information is an essential task, because processing and storing noisy information produce a waste of resources such as bandwidth, storage space, and computing time. Besides, the detection and extraction of the main content is useful in different areas, such as data mining, web summarization, and content adaptation to low resolutions. This work introduces a new technique for main content extraction. In contrast to most techniques, this technique not only extracts text, but also other types of content, such as images, and animations. It is a Document Object Model-based page-level technique, thus it only needs to load one single webpage to extract the main content. As a consequence, it is efficient enough as to be used online (in real-time). We have empirically evaluated the technique using a suite of real heterogeneous benchmarks producing very good results compared with other well-known content extraction techniques., This work has been partially supported by the EU (FEDER) and the Spanish MCI/AEI under grants TIN2016-76843-C4-1-R and PID2019-104735RB-C41, by the Generalitat Valenciana under grant Prometeo/2019/098 (DeepTrust), and by TAILOR, a project funded by EU Horizon 2020 research and innovation programme under GA No 952215.

With the rapidly growing amount of information on the internet, real-time system is one of the key strategies to cope with the information overload and to help users in finding highly relevant information. Real-time events and domain-specific information are important knowledge base references on the Web that frequently accessed by millions of users. Real-time system is a vital to product and a technique must resolve the context of challenges to be more reliable, e.g. short data life-cycles, heterogeneous user interests, strict time constraints, and context-dependent article relevance. Since real-time data have only a short time to live, real-time models have to be continuously adapted, ensuring that real-time data are always up-to-date. The focal point of this manuscript is for designing a real-time web search approach that aggregates several web search algorithms at query time to tune search results for relevancy. We learn a context-aware delegation algorithm that allows choosing the best real-time algorithms for each query request. The evaluation showed that the proposed approach outperforms the traditional models, in which it allows us to adapt the specific properties of the considered real-time resources. In the experiments, we found that it is highly relevant for most recently searched queries, consistent in its performance, and resilient to the drawbacks faced by other algorithms

Fuzzy C-means (FCM) is an important clustering algorithm with broad applications such as retail market data analysis, network monitoring, web usage mining, and stock market prediction. Especially, parameters in FCM have influence on clustering results. However, a lot of FCM algorithm did not solve the problem, that is, how to set parameters. In this study, we present a kind of method for computing parameters values according to role of parameters in the clustering process. New parameters are assigned to membership and typicality so as to modify objective function, on the basis of which Lagrange equation is constructed and iterative equation of membership is acquired, so does the typicality and center equation. At last, a new possibilistic fuzzy C-means based on the weight parameter algorithm (WPFCM) was proposed. In order to test the efficiency of the algorithm, some experiments on different datasets are conducted to compare WPFCM with FCM, possibilistic C-means (PCM), and possibilistic fuzzy C-means (PFCM). Experimental results show that iterative times of WPFCM are less than FCM about 25% and PFCM about 65% on dataset X12. Resubstitution errors of WPFCM are less than FCM about 19% and PCM about 74% and PFCM about 10% on the IRIS dataset.

SimRank is a popular measurement for evaluating the node-to-node similarities based on the graph topology. In recent years, single-source and top-k SimRank queries have received increasing attention due to their applications in web mining, social network analysis, and spam detection. However, a fundamental obstacle in studying SimRank has been the lack of ground truths. The only exact algorithm, Power Method, is computationally infeasible on graphs with more than $$10^6$$ nodes. Consequently, no existing work has evaluated the actual accuracy of various single-source and top-k SimRank algorithms on large real-world graphs. In this paper, we present ExactSim, the first algorithm that computes the exact single-source and top-k SimRank results on large graphs. This algorithm produces ground truths with precision up to 7 decimal places with high probability. With the ground truths computed by ExactSim, we present the first experimental study of the accuracy/cost trade-offs of existing approximate SimRank algorithms on large real-world graphs and synthetic graphs. Finally, we use the ground truths to exploit various properties of SimRank distributions on large graphs.

This research provides a review of the state of the art with respect to EDM and discusses the most relevant work in this area to date. Each study has been discussed considering type of data and data mining techniques used, and the kind of the educational task that they resolve. EDM is upcoming research area related to well-established areas of research such as e- learning, tutoring systems, web mining, data mining. Current literature show how fast educational data analysis area is growing and there is an increasing number of contributions that publish in International Journals and Conferences every year. However, educational data mining is still not a mature area. Some interesting future suggestion to develop this area has been presented. This research is a presentation of current and ancient literature of Predicting Student Performance using Data Mining.

Purpose This study aims to provide a system capable of static searching on a large number of unstructured texts directly on the Web domain while keeping costs to a minimum. The proposed framework is applied to the unstructured texts of Migne’s Patrologia Graeca (PG) collection, setting PG as an implementation example of the method. Design/methodology/approach The unstructured texts of PG have automatically transformed to a read-only not only Structured Query Language (NoSQL) database with a structure identical to that of a representational state transfer access point interface. The transformation makes it possible to execute queries and retrieve ranked results based on a specialized application of the extended Boolean model. Findings Using a specifically built Web-browser-based search tool, the user can quickly locate ranked relevant fragments of texts with the ability to navigate back and forth. The user can search using the initial part of words and by ignoring the diacritics of the Greek language. The performance of the search system is comparatively examined when different versions of hypertext transfer protocol (Http) are used for various network latencies and different modes of network connections. Queries using Http-2 have by far the best performance, compared to any of Http-1.1 modes. Originality/value The system is not limited to the case study of PG and has a generic application in the field of humanities. The expandability of the system in terms of semantic enrichment is feasible by taking into account synonyms and topics if they are available. The system’s main advantage is that it is totally static which implies important features such as simplicity, efficiency, fast response, portability, security and scalability.

How to integrate multimodal information sources for face naming in multimodal news is a hot and yet challenging problem. A novel deep cross-modal face naming scheme is developed in this paper to facilitate more effective people news retrieval for large-scale multimodal news. This scheme integrates deep multimodal analysis, cross-modal correlation learning, and multimodal information mining, in which the efficient naming mechanism aims to cluster the deep features of different modalities into a common space to explore their inter-related correlations, and a special Web mining pattern is designed to optimize the name-face matching for rare non-celebrity. Such a cross-modal face naming model can be treated as a problem of bi-media semantic mapping and modeled as an inter-related correlation distribution over deep representations of multimodal news, in which the most important is to create more effective cross-modal name-face correlation and measure to what degree they are correlated. The experiments on a large number of public data from Yahoo! News have obtained very positive results and demonstrated the effectiveness of the proposed model.

Data preprocessing is an essential task to prepare suitable target datasets to apply statistical and data mining algorithms. It has become one of the complex segment in Web usage mining due to the ...

The task of fuzzy clustering data is very interesting and important problem and often found Data Mining, Dynamic Data Mining, Data Stream Mining, Big Data Mining, Web Mining, Text Mining, etc. One of the main areas of computational intelligence are evolutionary algorithms that essentially represent certain mathematical models of biological organisms evolution. The goal of the paper is to propose the procedure of the adaptive methods of credibilistic fuzzy clustering using evolutionary algorithm. The goal of the work is adaptive credibilistic fuzzy clustering of data, using of credibility theory and evolutionary algorithm. The procedure of adaptive credibilistic fuzzy clustering of data based on the used goal function of special type and evolutionary approach from cat swarm optimization algorithm. This method designed to work both in batch and online mode, when data are fed to processing sequentially in real time. Proposed approach characterized by simple numerical implementation and relatively high rate of convergence. Proposed optimization method as a representative of evolutionary algorithms is intended for use in hybrid systems of computational intelligence, and especially in the problems of learning artificial neural networks, neuro-phase systems, as well as in the problems of clustering and classification. The experement result has confermed effectiveness stability work of adaptive credibilistic fuzzy clustering method based on evolutionary algorithm. Proposed method of adaptive credibilistic fuzzy clustering method based on evolutionary algorithm designed for use in hybrid systems of computational intelligence, in the problems of learning artificial neural networks, neuro-fuzzy systems, clustering and classification.

In the recent years with the advancement in technology, a lot of information is available in different formats and extracting the knowledge from that data has become a very difficult task. Due to the vast amount of information available on the web, users are finding it difficult to extract relevant information or create new knowledge using information available on the web. To solve this problem Web mining techniques are used to discover the interesting patterns from the hidden data .Web Usage Mining (WUM), which is one of the subset of Web Mining helps in extracting the hidden knowledge present in the Web log files , in recognizing various interests of web users and also in discovering customer behaviours. Web Usage mining includes different phases of data mining techniques called Data Pre-processing, Pattern Discovery & Pattern Analysis. This paper presents an updated focused survey on various sequential pattern mining algorithms like apriori-based algorithm , Breadth First Search-based strategy, Depth First Search strategy, sequential closed-pattern algorithm and Incremental pattern mining algorithm which are used in Pattern Discovery Phase of WUM. At last , a comparison is done based on the important key features present in these algorithms. This study gives us better understanding of the approaches of sequential pattern mining.

Purpose This paper aims to present a new term weighting approach for text classification as a text mining task. The original method, neutrosophic term frequency – inverse term frequency (NTF-IDF), is an extended version of the popular fuzzy TF-IDF (FTF-IDF) and uses the neutrosophic reasoning to analyze and generate weights for terms in natural languages. The paper also propose a comparative study between the popular FTF-IDF and NTF-IDF and their impacts on different machine learning (ML) classifiers for document categorization goals. Design/methodology/approach After preprocessing textual data, the original Neutrosophic TF-IDF applies the neutrosophic inference system (NIS) to produce weights for terms representing a document. Using the local frequency TF, global frequency IDF and text N's length as NIS inputs, this study generate two neutrosophic weights for a given term. The first measure provides information on the relevance degree for a word, and the second one represents their ambiguity degree. Next, the Zhang combination function is applied to combine neutrosophic weights outputs and present the final term weight, inserted in the document's representative vector. To analyze the NTF-IDF impact on the classification phase, this study uses a set of ML algorithms. Findings Practicing the neutrosophic logic (NL) characteristics, the authors have been able to study the ambiguity of the terms and their degree of relevance to represent a document. NL's choice has proven its effectiveness in defining significant text vectorization weights, especially for text classification tasks. The experimentation part demonstrates that the new method positively impacts the categorization. Moreover, the adopted system's recognition rate is higher than 91%, an accuracy score not attained using the FTF-IDF. Also, using benchmarked data sets, in different text mining fields, and many ML classifiers, i.e. SVM and Feed-Forward Network, and applying the proposed term scores NTF-IDF improves the accuracy by 10%. Originality/value The novelty of this paper lies in two aspects. First, a new term weighting method, which uses the term frequencies as components to define the relevance and the ambiguity of term; second, the application of NL to infer weights is considered as an original model in this paper, which also aims to correct the shortcomings of the FTF-IDF which uses fuzzy logic and its drawbacks. The introduced technique was combined with different ML models to improve the accuracy and relevance of the obtained feature vectors to fed the classification mechanism.

The problem of finding relevant data while searching the internet represents a big challenge for web users due to the enormous amounts of available information on the web. These difficulties are related to the well-known problem of information overload. In this work, we propose an online web assistant called OWNA. We developed a fully integrated framework for making recommendations in real-time based on web usage mining techniques. Our work starts with preparing raw data, then extracting useful information that helps build a knowledge base as well as assigns a specific weight for certain factors. The experiments show the advantages of the proposed model against alternative approaches.

E-Commerce berkembang pesat dalam world wide web hingga menghasilkan berbagai jenis data yang dapat dianalisa lebih lanjut untuk berbagai keperluan seperti personifikasi web, profiling customer, dan sebagainya. Salah satu jenis data yang dihasilkan e-Commerce adalah click stream data web yang merekam aktivitas visitor web dalam bentuk log data selama berinteraksi pada laman web. Penelitian ini mengekstraksi click stream data web e-commerce untuk mendapatkan pola interaksi konsumen terhadap halaman web selama mengunjungi web e-commerce. Berdasarkan jenis data yang diekstrak maka web usage mining digunakan untuk ekstraksi pola dari click stream data yang berbentuk log data. Teknik mining yang dianalisa terhadap log data e-commerce pada penelitian ini terdiri dari frequent itemset, asociation rules, dan frequence sequence mining. Frequent itemset menghasilkan halaman web yang paling sering diakses oleh visitor. Association rules menghasilkan pola kemungkinan halaman web yang akan diakses visitor jika visitor mengakses halaman-halamn tertentu. Frequence sequence mining mendapatkan pola urutan halaman web yang paling sering diakses oleh visitor web e-commerce saat berinteraksi pada laman web. Pola urutan halaman yang diakses visitor menunjukkan urutan kebiasaan visitor mengunjungi e-commerce. Sedangkan teknik mining yang diimplementasikan untuk menghasilkan pola akses visitor pada penelitian ini adalah Frequence sequence mining. Hasil ekstraksi dari penelitian ini menunjukkan ada enam halaman web yang paling sering diakses oleh konsumen dengan berbagai pola urutan aksesnya.

Lead users are the group of most valuable users in product innovation and new product development for a certain firm. Based on web mining methods on user innovation communities, this paper presents quantitative methodology to evaluate the contributions and values of online community users to identify lead users. Firstly, we analyze user behaviors and calculate the user interest index (UII) of posts, keywords and innovation fields to measure the popularity of innovations based on other users' focus. Secondly, the model of User Innovation Knowledge supernetwork (UIKSN) is proposed, in which UIIs and user contributions are considered as two types of node weights for integrating behavior data and content data. And then, rules and methods are suggested based on the UIKSN model for identifying lead users to meet various requirements including contributions and UIIs in posts, keywords, hot frontiers, core innovation fields, and even in certain fields or knowledge points. Furthermore, knowledge structures are analyzed through ego-network analysis. Case studies show that the proposed UIKSN model and methodology are more objective and credible and thus have good potentials for identifying and analyzing lead users from multiple levels, such as posts, keywords, hot front, core innovation fields, etc.

We investigate the degree and duration of the effects of deceptive signaling on globalized digital platforms, vis-a-vis naturally formed organic signaling such as importer review scores and discernible inducing signaling such as banner advertisements. Sponsored listing has emerged as a powerful tool for facilitating Internet-mediated exchanges in internationalization activities. Some digital platforms based in weakly regulated countries provide sponsored listing to exporters without explicitly disclosing that sponsorship. This constitutes a deceptive signal to importers in strongly regulated countries whose governments require sponsorship disclosure. Deceptive signaling is hypothesized to stimulate an institutional hypnosis on importers, with the most immediate and largest enhancing effects on exporter performance of the three types of signaling. Organic signaling is hypothesized to exert the most durable effect. Using transaction-level data from a globalized digital platform assembled with web mining methods, we develop a novel algorithm to detect deceptive signaling. Regression results from generalized method of moments estimation support these hypotheses. Our study is among the first to investigate institutional hypnosis manifested in increasingly widespread and unchecked forms in digitized international business transacting. By discovering and reporting on a dark side of institutional arbitrage, we offer actionable implications for digital platform governance, participants, and regulators.

Purpose The purpose of this paper is to define a methodology to analyze links between programming topics and libraries starting from GitHub data. Design/methodology/approach This paper developed an analysis over machine learning repositories on GitHub, finding communities of repositories and studying the anatomy of collaboration around a popular topic such as machine learning. Findings This analysis indicates the significant importance of programming languages and technologies such as Python and Jupyter Notebook. It also shows the rise of deep learning and of specific libraries such as Tensorflow from Google. Originality/value There exists no survey or analysis based on how developers influence each other for specific topics. Other researchers focused their analysis on the collaborative structure and social impact instead of topic impact. Using this methodology to analyze programming topics is important not just for machine learning but also for other topics.

Web siteleri, kurumsal ya da bireysel açıdan hitap edilen kitleyle ilk temasın sağlandığı bir etkileşim aracıdır. Bu araç, yoğun bir bilgi erişim ve ziyaret trafiğinin bulunduğu süreçlerde, kullanıcı davranışlarındaki farklı desenlerin tespit edilebileceği önemli bir potansiyeli içinde barındırmaktadır. Bu desenler, kullanıcı ihtiyaçlarının daha belirginleştirilmesi ve site geliştiricilerinin bu ihtiyaçlar doğrultusunda güncellemeler yapabilmesi açısından oldukça kritik görevler üstlenebilir. Bu çalışmanın amacı, dünya genelinde yaşanan Covid-19 pandemisinin ülkemizde etkinliğini arttırdığı süreçte, Kırklareli Üniversitesi web sitelerindeki kullanıcıların bilgi erişim ihtiyaçlarındaki değişimin belirlenmesidir. Bu amaç doğrultusunda, kullanıcıların bilgi erişim ve ziyaret davranışları, apriori algoritmasıyla bağımsız ve birlikte olacak şekilde incelenerek, aralarındaki ilişkilerin ortaya çıkarılması hedeflenmiştir. Bilgi erişim kavramı açısından çalışma sonuçları, kullanıcıların “tez yazımı”na yönelik çeşitli arama terimleriyle bilgi ihtiyaçlarını karşılamaya çalıştıklarını göstermiştir. Bu sonuç, özellikle lisansüstü öğrencilerin ilgili süreçte aktif olduklarına işaret etmektedir. Ziyaret davranışları açısından, “uzaktan eğitim”, “koronavirüs” ve “tatil” temalı sayfaların ağırlıklı olarak ziyaret edildiği ortaya çıkmıştır. Bilgi erişim davranışları sonrasında sergilenen ziyaret davranışları açısındansa, “tez yazımı”, “tatil” ve “eğitim öğretimin ertelenmesi” temalı ziyaretlerin birliktelikleri göze çarpmıştır. Çalışma sonucunda ortaya çıkarılmış olan davranış desenleri ve bu desenlerden nasıl faydalanılabileceğine yönelik öneriler çalışma kapsamında detaylı bir şekilde açıklanmıştır.

The rapid increase in web sites has created large volumes of data in web environment, which generates a problem of how to extract and gain useful knowledge from such data. Extracting such knowledge helps to discover, understand and predict user behaviors based on his interaction with the website. This paper introduces two neural network based systems, one for web visitor recognition according to their web logs pattern, and the other for web visitor classification according to the visited pages. This will introduce rapid services and save user time with web through a database connected to these neural networks Such system can be used to improve efficiency and effectiveness in searching for information on the web. Complete architecture of the networks is given based on supervised and unsupervised learning paradigms. Experiments have been carried out in order to validate this approach.

The comprehension of web visitors patterns are needed to develop website in an optimal fashion. The visitor pattern contained in the web log file of Diponegoro University’s website is clustered by Modified Gustafson-Kessel method. In general, this method produces two until six clusters. Two kinds of results are outlined in this paper. The first is the result contains two clusters, and the second is containing three clusters. In the first result, the visitors are divided into information seekers of student capacity and Engineering Faculty. In the second result, the visitors are divided into information seekers of Medicine Faculty, student admission and Engineering Faculty.

This study explores the use of web content analysis to build innovation indicators from the complete texts of 79 corporate websites of Canadian nanotechnology and advanced materials firms. Indicators of four core concepts (R&D, IP protection, collaboration, and external financing) of the innovation process were built using keywords frequency analysis. These web-based indicators were validated using several indicators built from a classic questionnaire-based survey with the following methods: correlation analysis, multitraits multimethods (MTMM) matrices, and confirmatory factor analysis (CFA). The results suggest that formative indices built with the questionnaire and web-based indicators measure the same concept, which is not the case when considering the items from the questionnaire separately. Web-based indicators can act either as complements to direct measures or as substitutes for broader measures, notably the importance of R&D and the importance of IP protection, which are normally measured using conventional methods, such as government administrative data or questionnaire-based surveys.

Existing approaches to model innovation ecosystems have been mostly restricted to qualitative and small-scale levels or, when relying on traditional innovation indicators such as patents and questionnaire-based survey, suffered from a lack of timeliness, granularity, and coverage. Websites of firms are a particularly interesting data source for innovation research, as they are used for publishing information about potentially innovative products, services, and cooperation with other firms. Analyzing the textual and relational content on these websites and extracting innovation-related information from them has the potential to provide researchers and policy-makers with a cost-effective way to survey millions of businesses and gain insights into their innovation activity, their cooperation, and applied technologies. For this purpose, we propose a web mining framework for consistent and reproducible mapping of innovation ecosystems. In a large-scale pilot study we use a database with 2.4 million German firms to test our framework and explore firm websites as a data source. Thereby we put particular emphasis on the investigation of a potential bias when surveying innovation systems through firm websites if only certain firm types can be surveyed using our proposed approach. We find that the availability of a websites and the characteristics of the website (number of subpages and hyperlinks, text volume, language used) differs according to firm size, age, location, and sector. We also find that patenting firms will be overrepresented in web mining studies. Web mining as a survey method also has to cope with extremely large and hyper-connected outlier websites and the fact that low broadband availability appears to prevent some firms from operating their own website and thus excludes them from web mining analysis. We then apply the proposed framework to map an exemplary innovation ecosystem of Berlin-based firms that are engaged in artificial intelligence. Finally, we outline several approaches how to transfer firm website content into valuable innovation indicators.

Data mining, which has different uses such as text mining and web mining, is especially used for clustering and classification purposes. In this study, this method was used for both classification and text mining. The aim of the study was the assessment of the performances of the data mining algorithms on the three datasets. A total of 6631 master's and doctoral dissertations written in the field of industrial engineering were downloaded from the Higher Education Council database. With the help of summary, subject titles and keywords of these dissertations, it was tried to be guessed which sub-field of industrial engineering it belongs to using WEKA program. As a result, it was observed that the data set containing the keywords obtained by weighting the expert opinion was more successful than the other two data sets. And the three most successful classification algorithms were found to be kNN, SMO, and J48, respectively. Keywords: Classification Algorithms, Data Mining, Multiple Classes, Dataset.

The purpose of the workis to explore the current problems and prospects of mining solution, big web data in real time, as well as the possibility of practical implementation of Web Mining technology for big web data on a practical example.Materials and methods.The study included a review of bibliographic sources on big data mining. We used Web Mining technology for associative analysis of large web data, as well as computer modeling of the practical task of transaction analysis using a general-purpose scripting language (PHP).Results.During the work, the specifics of the Data Mining technology were described, and a modern approach to the analysis of large web data –Web Mining was analyzed. A brief classification of tasks solved using Web Mining technology is given. The problem of data mining of large web data in a general-purpose scripting language (PHP) has been solved: the lack of libraries for data mining, the difficult normalization of data to the form necessary for associative analysis, interaction with the database management system. Also, an example showing an approach to the mining of large web data was implemented. Based on the understanding of Web Mining technology and the described difficulties of analyzing web data in the PHP language, methods for effectively solving the practical problem of analyzing web data based on transactions committed in a dynamic web application have been proposed. A module for associative analysis of customer transactions in the programming language PHP was developed. The module includes an intelligent data processing class. The structural scheme of the module and system architecture were developed. The constructed module allows us to solve the main part of the problem of associative analysis of large web data using Web Mining technology in order to solve the problem of identifying patterns in a large array of web data. Associative analysis of web data is much faster because of the combination of a general-purpose scripting language and an object-oriented approach.Conclusion.According to the results of the study, it can be argued that the current state of the technology for the analysis of large web data allows efficiently process data objects, identify patterns, obtain hidden data and receive complete statistical data in real time. The results can be used both for the purpose of the initial research of technologies for analyzing large web data, and as an addition to the content management system for the intelligent analysis of web data. The usage of the technology of associative analysis and the created universal handler class makes the created module flexible, while the possibility of manual integration makes this module universal. With manual integration, the database management system is not important. Algorithm methods work with selected data. This factor greatly simplifies the further development of program code.

Purpose This study aims to analyze Kickstarter data along with social media data from a data mining perspective. Kickstarter is a crowdfunding financing plataform and is a form of fundraising and is increasingly being adopted as a source for achieving the viability of projects. Despite its importance and adoption growth, the success rate of crowdfunding campaigns was 47% in 2017, and it has decreased over the years. A way of increasing the chances of success of campaigns would be to predict, by using machine learning techniques, if a campaign would be successful. By applying classification models, it is possible to estimate if whether or not a campaign will achieve success, and by applying regression models, the authors can forecast the amount of money to be funded. Design/methodology/approach The authors propose a solution in two phases, namely, launching and campaigning. As a result, models better suited for each point in time of a campaign life cycle. Findings The authors produced a static predictor capable of classifying the campaigns with an accuracy of 71%. The regression method for phase one achieved a 6.45 of root mean squared error. The dynamic classifier was able to achieve 85% of accuracy before 10% of campaign duration, the equivalent of 3 days, given a campaign with 30 days of length. At this same period time, it was able to achieve a forecasting performance of 2.5 of root mean squared error. Originality/value The authors carry out this research presenting the results with a set of real data from a crowdfunding platform. The results are discussed according to the existing literature. This provides a comprehensive review, detailing important research instructions for advancing this field of literature.

PurposeThis study intends to examine consumers' fashion customization experiences through a web content mining (WCM) approach. By applying the theory of customer value, this study explores the benefits and costs of two levels of mass customization (MC) to identify the values derived from style (i.e. shoe customization) and fit customization experiences (i.e. apparel customization) and further to compare the dominating dimensions of value derived across style and fit customization.Design/methodology/approachA WCM approach was applied. Also, two case studies were conducted with one focusing on style customization and the other focusing on fit customization. The brand Vans was selected to examine style customization in study 1. The brand Sumissura was selected to examine fit customization in study 2. Consumers' comments on customization experiences from these two brands were collected through social networks, respectively. After data cleaning, 394 reviews for Vans and 510 reviews for Sumissura were included in the final data analysis. Co-occurrence plots, feature extraction and grouping were used for the data analysis.FindingsThe emotional value was found to be the major benefit for style customization, while the functional value was indicated as the major benefit for fit customization, followed by ease of use and emotional value. In addition, three major themes of costs, including unsatisfied service, disappointing product performance and financial risk, were revealed by excavating and evaluating consumers' feedback of their actual clothing customization experiences with Sumissura.Originality/valueThis study initiates the effort to use web mining, specifically, the WCM approach to thoroughly investigate the benefits and costs of MC through real consumers' feedback of two different types of fashion products. The analysis of this study also reflects the levels of customization: style and fit. It provides an in-depth text analysis of online MC consumers' feedback through the use of feature extraction analysis and word co-occurrence networks.

Transfer learning is focused on the reuse of supervised learning models in a new context. Prominent applications can be found in robotics, image processing or web mining. In these fields, the learning scenarios are naturally changing but often remain related to each other motivating the reuse of existing supervised models. Current transfer learning models are neither sparse nor interpretable. Sparsity is very desirable if the methods have to be used in technically limited environments and interpretability is getting more critical due to privacy regulations. In this work, we propose two transfer learning extensions integrated into the sparse and interpretable probabilistic classification vector machine. They are compared to standard benchmarks in the field and show their relevance either by sparsity or performance improvements., arXiv admin note: text overlap with arXiv:1907.01343

Web Usage Mining applies fewer techniques in record data to pull out the behavior of users. The knowledge mined from the web log can be utilized in web personalization, Prediction, prefetching, restructuring of web sites etc. It consists of three steps in preprocessing, pattern detection and analysis. Web log information is typically noisy and uncertain and preprocessing is a significant process ahead of mining. The Patterns discovered after applying the mining techniques are dependent on the accuracy of the weblog which in turn depends on the preprocessing phase. The output of preprocessing should be the user’s navigation session file. In this paper the techniques of preprocessing and the method for construction of user’s navigation session file is proposed.

Purpose This study aims to highlight the challenges and opportunities of using GitHub as a data source in both research and programming education. Design/methodology/approach This study provides general overview of the challenges and opportunities faced while conducting empirical research using GitHub as a data source. The challenges and opportunities are framed using the input–process–output model of open-source software. Findings GitHub data accessed from the application programming interface (API) can have several limitations, which can be overcome by Web scraping and using external data repositories such as GHArchive and GHTorrent. There are also several idiosyncrasies about GitHub that researchers need to be aware of to be able to use the data effectively, which can represent an opportunity for research. The challenges and opportunities are summarized for the licenses, community, development process and product of free/libra and open-source software communities hosted on GitHub. Originality/value This study provides a summary of GitHub-related challenges and opportunities that researchers can leverage to improve their empirical research. Furthermore, this summary can be a valuable resource for instructors that plan to use GitHub as a data source in their data-focused programming courses.

PurposeStrategic alliances among organizations are some of the central drivers of innovation and economic growth. However, the discovery of alliances has relied on pure manual search and has limited scope. This paper proposes a text-mining framework, ACRank, that automatically extracts alliances from news articles. ACRank aims to provide human analysts with a higher coverage of strategic alliances compared to existing databases, yet maintain a reasonable extraction precision. It has the potential to discover alliances involving less well-known companies, a situation often neglected by commercial databases.Design/methodology/approachThe proposed framework is a systematic process of alliance extraction and validation using natural language processing techniques and alliance domain knowledge. The process integrates news article search, entity extraction, and syntactic and semantic linguistic parsing techniques. In particular, Alliance Discovery Template (ADT) identifies a number of linguistic templates expanded from expert domain knowledge and extract potential alliances at sentence-level. Alliance Confidence Ranking (ACRank)further validates each unique alliance based on multiple features at document-level. The framework is designed to deal with extremely skewed, noisy data from news articles.FindingsIn evaluating the performance of ACRank on a gold standard data set of IBM alliances (2006–2008) showed that: Sentence-level ADT-based extraction achieved 78.1% recall and 44.7% precision and eliminated over 99% of the noise in news articles. ACRank further improved precision to 97% with the top20% of extracted alliance instances. Further comparison with Thomson Reuters SDC database showed that SDC covered less than 20% of total alliances, while ACRank covered 67%. When applying ACRank to Dow 30 company news articles, ACRank is estimated to achieve a recall between 0.48 and 0.95, and only 15% of the alliances appeared in SDC.Originality/valueThe research framework proposed in this paper indicates a promising direction of building a comprehensive alliance database using automatic approaches. It adds value to academic studies and business analyses that require in-depth knowledge of strategic alliances. It also encourages other innovative studies that use text mining and data analytics to study business relations.

Web usage data usually contain quantitative values, and this implies that fuzzy logic can be used to represent such values. The time spent by users on each web page is a part of web usage data, which can be used to analyze users’ browsing behavior. In existing research on fuzzy web mining, the time duration of web pages is shown as trapezoidal membership functions (TMFs), and the number and parameters of TMFs are already predefined. TMFs of each web page are different from those of other web pages. Therefore, instead of using predefined TMFs, in this study, we proposed a new algorithm called CALA-FOMF to find both the number of TMFs and their optimized parameters to mine fuzzy association rules in web usage data using a team of continuous action-set learning automata (CALA). CALA-FOMF contained two steps. In the first step, using a team of CALA, we introduced a new framework. The proposed framework obtained the number of TMFs as inputs and found their optimized parameters. The proposed framework was able to reduce the search space and eliminate inappropriate membership functions during the learning process. In the second step, we proposed a new algorithm using the proposed framework to find an appropriate number of TMFs and their optimized parameters. The performance of the CALA-FOMF approach was compared with that of the fuzzy web mining algorithm, which used uniform TMFs. Experiments on datasets with different sizes confirmed that the proposed CALA-FOMF increased the efficiency of mining fuzzy association rules by extracting optimized TMFs.

Thanks to recent efforts by the text mining community, biocurators have now access to plenty of good tools and Web interfaces for identifying and visualizing biomedical entities in literature. Yet, many of these systems start with a PubMed query, which is limited by strong Boolean constraints. Some semantic search engines exploit entities for Information Retrieval, and/or deliver relevance-based ranked results. Yet, they are not designed for supporting a specific curation workflow, and allow very limited control on the search process. The Swiss Institute of Bioinformatics Literature Services (SIBiLS) provide personalized Information Retrieval in the biological literature. Indeed, SIBiLS allow fully customizable search in semantically enriched contents, based on keywords and/or mapped biomedical entities from a growing set of standardized and legacy vocabularies. The services have been used and favourably evaluated to assist the curation of genes and gene products, by delivering customized literature triage engines to different curation teams. SIBiLS (https://candy.hesge.ch/SIBiLS) are freely accessible via REST APIs and are ready to empower any curation workflow, built on modern technologies scalable with big data: MongoDB and Elasticsearch. They cover MEDLINE and PubMed Central Open Access enriched by nearly 2 billion of mapped biomedical entities, and are daily updated.

More than two decades, there is a number of weather-related websites are available which approximately predict the weather and climate. By extracting essential data from the websites, a predictive data pattern can be produced to show the next day’s weather is with rain or not. By applying different types of web mining and analyzing techniques those extracted weather-related data can be visualized to a typical pattern for weather forecasting with the main deciding factors of weather. With the use of these approaches, reasonably precise forecasts can be made up to about four to five days in advance. For the weather prediction analysis, we need to discover deciding factors of the next day’s weather. Particularly, common weather dependent factors and the relationship of the prediction to the particular phenomenon. The solution proposed by this research can be used to analyze a large amount of weather data which are in different forms in each source. By using predictive mining task our solution allows us to make predictions for future instances according to the model what we have created. Evaluation measurements for the selected data mining technique such as accuracy percentage, TP & FP Rate, Precision, F-Measure, ROC area, SSE, and loglikelihood for classification and clustering leads to create a high-quality model of prediction. Knowledge flow interface provides the data flow to show the processing and analyzing data with precise association rules. In order to evaluate the model, SSE values and time to build the model, are considered in an effective manner.

The web today has become an archive of information in any structure such content, sound, video, designs, and multimedia, with the progression of time overall web, the world wide web is now crowded with different data making extraction of virtual data burdensome process, web utilizes various information mining strategies to mine helpful information from page substance and web hyperlink. The fundamental employments of web content mining are to gather, sort out, classify, providing the best data accessible on the web for the client who needs to get it. The WCM tools are needful to examining some HTML reports, content and pictures at that point, the outcome is using by the web engine. This paper displays an overview of web mining categorization, web content technique and critical review and study of web content mining tools since (2011-2019) by building the table's a comparison of these instruments dependent on some important criteria

In recent years, the use of information technologies in Critical Infrastructures is gradually increasing. Although this brings benefits, it also increases the possibility of security attacks. Despite the availability of various advanced incident handling techniques and tools, there is still no easy, structured, standardized and trusted way to manage and forecast interrelated cybersecurity incidents. This paper introduces CyberSANE, a novel dynamic and collaborative, warning and response system, which supports security officers and operators to recognize, identify, dynamically analyse, forecast, treat and respond to security threats and risks and and it guides them to handle effectively cyber incidents. The components of CyberSANE are described along with a description of the CyberSANE data flow. The main novelty of the CyberSANE system is the fact that it enables the combination of active incident handling approaches with reactive approaches to support incidents of compound, highly dependent Critical Information Infrastructures. The benefits and added value of using CyberSANE is described with the aid of a set of cyber-attack scenarios.

With the evolution in technology huge amount of data is being generated, and extracts the necessary data from large volumes of data. This process is significantly complex. Generally the web contains bulk of raw data and the process of converting this data to information mining process can be performed. At whatever point the user places some inquiry on particular web search tool, outcomes are produced with respect to the requests which are dependent on the magnitude of the document created via web information retrieval tools. The results are obtained using calculations and implementation of well written algorithms. Well known web search tools like Google and other varied engines contain their specific manner to compute the page rank, various outcomes are obtained on various web crawlers for a same inquiry because the method for deciding the importance of the sites contrasts among number of algorithm. In this research, an attempt to analyze well-known page ranking calculation on the basis of their quality and shortcomings. This paper places the light on a portion of the extremely mainstream ranking algorithm and attempts to discover a better arrangement that can optimize the time spent on looking through the list of sites.

E-commerce provides a global platform supporting product transactions through the consumer purchase lifecycle including communications of perceived satisfaction and dissatisfaction. The customer feedback functions and social networks of many e-commerce websites allow for the creation of extremely large databases that can be mined to model the customers' perceptions toward online purchases. This research uses online customer reviews as the business intelligence corpus to help companies redesign products that better satisfy consumer preferences and differentiate their product offerings. After identifying the specific webpages of customer reviews, a web crawler collects review text. Computer-supported text mining, cluster analysis, and perceptual mapping are combined as a systematic analytic approach to compare products in a given domain. The study assists phone manufacturers to understand the positive and negative perceptions of customers related to their post-purchase experiences. The customer-preferred product functions, features, and price positions provide valuable strategic intelligence for new product designs and market differentiation.