A Clustering Model for Mining Evolving Web User Patterns in Data Stream Environment

Significant effort is currently invested in application integration, enabling business processes of different companies to interact and form complex multiparty processes. Web service standards, based on WSDL (Web Service Definition Language), have been adopted as process-to-process communication paradigms. However, the conceptual modeling of applications using Web services has not yet been addressed. Interaction with Web services is often specified at the level of the source code; thus, Web service interfaces are buried within a programmatic specification.In this article, we argue that Web services should be considered first-class citizens in the specification of Web applications. Thus, service-enabled Web applications should benefit from the high-level modeling and automatic code generation techniques that have long been advocated for Web application design and implementation. To this end, we extend a declarative model for specifying data-intensive Web applications in two directions: (i) high-level modeling of Web services and their interactions with the Web applications which use them, and (ii) modeling and specification of Web applications implementing new, complex Web services.Our approach is fully implemented within a CASE tool allowing the high-level modeling and automatic deployment of service-enabled Web applications.

Due to problems in correctly understanding user requirements the Information System (IS) development community have recognised the need to involve end-users in the development and maintenance of web applications. End-users perceive web applications through user interfaces (UIs) and commonly use sketches of UIs to express their requirements. Thus, it would be desirable to provide an end-user development methodology centred on UI modelling techniques. In this paper a visual modelling approach is presented to empower end-users in developing data intensive web applications starting from user interface descriptions. The modelling language follows a holistic approach by representing both static and behavioural information of a web application in one visual model. The visual model allows the specification of the look-and-feel of the application through mock-ups, and the user interactions through links, annotations, and widget references. End-users are guided during the modelling process by providing a summary view to manage the design of complex applications and a data model view to improve the quality of the generated applications.

This paper addresses conceptual modeling and automatic code generation for search engine integration with data intensive Web applications. We have analyzed the similarities (and differences) between IR and database systems to extend an existing domain specific language for data-driven Web applications. The extended Web modeling language specifies the search engine’s index schemas based on the data schema of the Web application and uniquely designs the interaction between the database, the Web application, the search engine and users. We also provide an implementation of a CASE tool extension for visual modeling and code generation. Experimentation of the proposed approach has been successfully applied in the context of the COOPER project.

The goal of disseminating and manipulating spatial knowledge over the Internet, has led to the development of Web applications, known as Web Geographic Information Systems (Web GIS). Web GIS can be considered as a particular class of data-intensive Web applications, since they are mainly devoted to handle (spatial) information to and from the user. The success of WebML (Web Modeling Language) for designing traditional data-intensive web applications suggested use to extend this visual formalism to model relevant interaction and navigation operations typical of Web GIS. In the paper, we describe the proposed extension and provide an example of its application.

Current data-intensive Web applications, such as on-line trading, e-commerce, corporate portals and so on, are becoming more and more complex, both in terms of density of information and in terms of navigational paths. At this aim different techniques have been proposed in literature for optimizing the information to be shown to the user. In this paper we present a technique for tuning the amount of data presented to the user, directly inspired to the concept of Levels of Details (LoD), commonly used in computer graphics. Like in computer graphics the idea is to simplify the original model, without loosing the main characteristics of the objects to be shown. The approach is based on the application of LoD operators to the compositional and navigational structure of a Web application, expressed through an hypertextual model.KeywordsAccess ModalityAccess PathIncoming LinkDetail ConceptNavigational PathThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Recommender systems have emerged as an important part of the solution to the information overload problem facing today's Web users. Combining ideas and techniques from information filtering, user modeling, artificial intelligence, user interface design and human-computer interaction, recommender systems provide users with proactive suggestions that are tailored to meet their particular information needs and preferences. Indeed recommender systems have enjoyed considerable commercial success and continue to play an increasingly important role in many online services, from Amazon and Yahoo! to iTunes and Last.fm. It is our great pleasure to welcome you to the Second ACM International Conference on Recommender Systems (RecSys), which builds on a wonderful legacy of research workshops, the Recommenders Summer School held in Bilbao in 2006, and the First ACM International Conference on Recommender Systems held in Minneapolis in 2007. This conference brings together leaders in the field, from both research and practice, to explore the latest innovations, challenges, and opportunities for recommender systems technologies. The call for papers attracted 121 submissions, a more than 100% increase compared to 2007, with authors from nearly 30 different countries. After a detailed scientific review process the program committee selected 38 papers for publication in the proceedings and presentation at the conference, a 31% acceptance rate. Of these 38 papers, 20 full-length papers were accepted for oral presentation and 18 papers for poster presentation, of which 15 were full-length papers and 3 were short-papers. In addition, the conference program includes a keynote presentation by Yahoo!'s Andrei Broder; two panel sessions filled with industry leaders sharing experiences and raising challenges for the research community; and presentations of six early-stage recommender start-ups, each bidding to win a $100,000 prize from Strands, Inc. The conference also includes three tutorials, in which leaders of the field explore hot topics in more detail; and a doctoral symposium to provide an opportunity for doctoral students to explore and develop their research interests in an interdisciplinary workshop, under the guidance of a panel of distinguished research faculty.

Our work focuses on tackling the problem of efficiency and accuracy of web usage clustering for recommender systems. Accurate analysis and preprocessing of web usage data and efficient web usage clustering are the key factors that influence the development of clustering based implicit recommender system. We propose an analysis and preprocessing model to tackle the poor quality of web usage data. To address the problem of efficient web usage clustering, we propose a Particle Swarm Optimization (PSO) based clustering approach. Having shown our PSO based clustering performs well; we extend it for mining the usage behavior of web users. We select Java API (Application Programming Interface) documentation usage data as a case study for our recommender system.

Classifying Web users based on multi-dimensional features is one of the foundations of realizing personalized Web applications. It could be used for user classification model, users' multi-dimensional data analysis, potential user group discovery and personalized recommendation and so forth. In this paper, a variable granularity user classification algorithm based on Web users' multidimensional features is proposed. Given a user feature model, the algorithm will mine all common feature categories and find the relationships between them. A series of experiments are conducted to analyze the performances of this algorithm with different condition. The experimental results indicate that this algorithm has good performance and can be deployed in Web applications with massive Web users.

Recommender systems are an effective solution to the information overload problem, especially in the World Wide Web where we gather vast information from anonymous people around the world. Trust enhanced recommender system to be promising to overcome the cold-start and sparsity challenges of traditional recommender system as well as improving the accuracy of the recommendations. This arise a research focus about blending of the trust information and trust level prediction to the recommendation framework. From the past decade, numerous researches were done to adapt online social network trust (simply social trust) for many web applications, including e-commerce, P2P networks, multi-agent systems, recommendation systems, and service-oriented computing. Usually, online social trust prediction can be based on two mechanisms to acquire trust value: evaluating trustee on basis of truster / truster's neighbor trust experience information (local trust), otherwise evaluating trustee on the basis of the whole social network trust experience information as reputation (global trust). Here, we leverage social science theories to develop the trust models that enable the study of online social trust evolution. In this paper, we propose a matrix factorization based trust enhanced recommendation system which properly incorporates both local trust and global trust with diffusion of the social trust multi-aspects to improve the quality of recommendations for mitigating the data sparsity and the cold-start issues. Through experiments on the Epinions data set, we show that our model outperforms its standard trust enhanced counterparts with respect to accuracy on recommender systems.

Photo-sharing websites such as Flickr and Panoramio contain millions of geotagged images contributed by people from all over the world. Characteristics of these data pose new challenges in the domain of spatio-temporal analysis. In this paper, we define several different tasks related to analysis of attractive places, points of interest and comparison of behavioral patterns of different user communities on geotagged photo data. We perform analysis and comparison of temporal events, rankings of sightseeing places in a city, and study mobility of people using geotagged photos. We take a systematic approach to accomplish these tasks by applying scalable computational techniques, using statistical and data mining algorithms, combined with interactive geo-visualization. We provide exploratory visual analysis environment, which allows the analyst to detect spatial and temporal patterns and extract additional knowledge from large geotagged photo collections. We demonstrate our approach by applying the methods to several regions in the world.

AI has rapidly changed how web applications are created and delivered in recent years. Using AI brings the user benefits of better automated processes through intelligent self-service, personalization of user experiences/customers, predictive analysis, and now includes real-time decision making. Before AI, web applications were mostly static or rules-based with limited or no change in response to how users interact with them. The introduction of advanced learning technologies such as Machine Learning (ML), Deep Learning (DL), and Natural Language Processing (NLP) has enabled web applications to become "intelligent" platforms that can receive input from a large volume of data and make adjustments in performance to continuously improve Keywords:Artificial Intelligence (AI), Web Applications, Machine Learning (ML), Deep Learning, Natural Language Processing (NLP), Recommendation Systems, Chatbots, Predictive Analytics, Data Mining.

From the Publisher: Programming the Web with Visual Basic .NET is a comprehensive guide to building Web applications and services using Visual Basic .NET. It is designed especially for experienced Visual Basic programmers who use Visual Studio .NET for their development work. The authors do not assume that readers know how to program Web applications. Because the .NET Framework simply blows away the archaic tools previously available to Web programmers, the authors predict that many Visual Basic programmers who successfully avoided Web programming in the past will now bring their expertise to the Web. Yet, experienced Web programmers will also benefit from the authors' thorough coverage of the ASP.NET namespaces and their clear coverage of the ADO.NET classes most important to Web applications that use relational databases for data storage. All developers will benefit from the authors' extensive practical advice (based on their unique professional backgrounds) about how to produce high quality code and how to design professional quality, usable Web pages. After reading Programming the Web with Visual Basic .NET, developers will understand how to build and deploy top quality, professionally designed, highly usable Web applications using Visual Basic .NET. The three co-authors of this book are related in both life and business. Constance Petersen and her husband, Lynn Torkelson, are the founders of SoftMedia Artisans, Inc. The third author is their fourteen-year-old son Zachary Torkelson. Lynn, a grizzled veteran of assembly language and C programming, has extensive experience using Visual Basic to develop client-server business applications. Before ditching the corporate world to go into business with Constance, Lynn also did tours of duty as a software development manager and as a manager of software quality. Constance co-authored the best-selling Visual Basic 4 How-To and has been recognized with the Microsoft Most Valuable Professional (MVP) award for her contributions to the VB development community. The creative force within SoftMedia Artisans, she has written extensively on Web design and usability. Zac began investigating the joys of programming at the age of ten and gradually intensified his interest. He soon learned VB5, then VB6, and C++. After discovering the works of Dan Appleman on the company shelves, Zac began using the Windows API extensively in his programs. Upon the arrival of the first .NET Beta, Zac switched to Visual Basic .NET and C#. These days he particularly likes to program multiplayer online games with cool graphics and sound. - Authors speak directly to experienced Visual Basic programmers who want to use Visual Basic .NET to program Web applications - Provides thorough coverage of the ASP.NET namespaces - Includes much practical advice about how to produce high quality code and how to design professional quality, usable Web pages—thanks to Authors' unique professional backgrounds ? Authors will maintain an ongoing Web site in support of the book (updated source code, errata, additional tips, articles, examples, etc.)

The tremendous growth in volume of web usage data results in the boost of web mining research with focus on discovering potentially useful knowledge from web usage data. This paper presents a new web usage mining process for finding sequential patterns in web usage data which can be used for predicting the possible next move in browsing sessions for web personalization. This process consists of three main stages: preprocessing web access sequences from the web server log, mining preprocessed web log access sequences by a tree-based algorithm, and predicting web access sequences by using a dynamic clustering-based model. It is designed based on the integration of the dynamic clustering-based Markov model with the Pre-Order Linked WAP-Tree Mining (PLWAP) algorithm to enhance mining performance. The proposed mining process is verified by experiments with promising results.

In recent years, it has been witnessed that the ever-interesting and upcoming publishing medium is the World Wide Web. Much of the web content is unstructured so gathering and making sense of such data is very tedious. Web servers worldwide generate a vast amount of information on web users' browsing activities. Several researchers have studied these so-called web access log data to better understand and characterize web users. Data can be enriched with information about the content of visited pages and the origin (e.g., geographic, organizational) of the requests. The goal of this project is to analyze user behavior by mining enriched web access log data. The several web usage mining methods for extracting useful features is discussed and employ all these techniques to cluster the users of the domain to study their behaviors comprehensively. The contributions of this thesis are a data enrichment that is content and origin based and a treelike visualization of frequent navigational sequences. This visualization allows for an easily interpretable tree-like view of patterns with highlighted relevant information. The results of this project can be applied on diverse purposes, including marketing, web content advising, (re-)structuring of web sites and several other E-business processes, like recommendation and advertiser systems. It also rank the best relevant documents based on Top K query for effective and efficient data retrieval system. It filters the web documents by providing the relevant content in the search engine result page (SERP).

Web extensions are powerful software artifacts that allow end-users to adapt and enrich a website. These extensions run on the user's web browser as a single-user software that manipulates available third-party web contents. Many of them offer some collaborative features that depend on a web application. The need of two co-depending software artifacts (the web application as back-end and the web extensions as front-end) increases complexity, making the system harder to develop and maintain. In this paper we tackle this problem by proposing a P2P approach to build collaborative web extensions. The approach involves a middleware and a framework. On the one hand, the middleware serves to manage the resources offered by the browser so multiple P2P extensions can coexist. It ensures that the overall performance of the browser is not degraded by the collaborative web extension. On the other hand, the proposed framework is intended to allow developers without experience in P2P to create collaborative web extensions on top of the middleware. This paper discusses the main challenges of building P2P web extensions, presents the approach, and two case studies focused on the use of the framework for inexperienced developers.