Narrow Results

A technique for creating analysis models of a web application from a business model is proposed for easy and effective development. Moreover, a technique for generating test cases from the sequence diagrams for a web application is proposed. The use case diagram and web page list are generated from the business model which is depicted using the notations of the UML (Unified Modeling Language) activity diagram. The page diagram and logical/physical database models are created based on the web page list and extracted data fields. Test cases for the web application are generated from call messages (including self-call messages) of the UML sequence diagram. The effectiveness of these techniques is shown using a practical case study which is a development project of a web application for RMS (Research material Management System).

One of the main characteristics of most Web applications is their high dynamism. Once implemented and deployed for the first time, new functionalities are added to meet new or changed requirements. Some of these functionalities may appear on the Web in response to an unexpected event, or phenomena (such as a natural calamity) after which they are removed. Some others are activated periodically, to coincide with a particular date, or period of the year (such as, return to school, Christmas holidays, etc.). Implementing such volatile functionalities usually impacts on a number of aspects of a Web application, including content, navigation, presentation, business processes, and user operations. Their cyclic activation/deactivation, which requires repetitive changes in the application code, may be the cause of waste of effort and application quality deterioration, up to incorrect functioning.

In this paper, we present an approach to decouple the design and implementation of volatile functionalities from that of stable ones, i.e. the core functionalities of the application. The approach is instantiated in the context of the Object-Oriented Hypermedia Design Method (OOHDM), but its principles and related techniques are generally applicable to any other Web engineering method. We show how our approach enables the deployment and removal of these functionalities in a cost-effective and safe way and at runtime, thus providing business agility. A framework to classify volatile functionalities and a number of examples are also reported.

The technology that integrates various types of Web contents to build a new Web application through end-user programming is widely used nowadays. However, the Web contents do not have a uniform interface for accessing the data and computation. Most of the general Web users access information on the Web through applications until now. Hence, designing a uniform and flexible programmatic interface for integration of different Web contents is unavoidable. In this paper, we propose an approach that can be used to analyze Web applications automatically and reuse the information of Web applications through the programmatic interface we designed. Our approach can support the flexible integration of Web applications, Web services and Web feeds. In our experiments, we use a large number of Web pages from different types of Web applications and achieve the integration by the proposed programmatic interfaces. The experimental results show that our approach brings to the end-users a flexible and user-friendly programming environment.

The software security issue is being paid great attention from the software development community as security violations have emerged variously. Developers often use access control techniques to restrict some security breaches to software systems’ resources. The addition of authorization constraints to the role-based access control model increases the ability to express access rules in real-world problems. However, the complexity of combining components, libraries and programming languages during the implementation stage of web systems’ access control policies may arise potential flaws that make applications’ access control policies inconsistent with their specifications. In this paper, we introduce an approach to review the implementation of these models in web applications written by Java EE according to the MVC architecture under the support of the Spring Security framework. The approach can help developers in detecting flaws in the assignment implementation process of the models. First, the approach focuses on extracting the information about users and roles from the database of the web application. We then analyze policy configuration files to establish the access analysis tree of the application. Next, algorithms are introduced to validate the correctness of the implemented user-role and role-permission assignments in the application system. Lastly, we developed a tool called VeRA, to automatically support the verification process. The tool is also experimented with a number of access violation scenarios in the medical record management system.

Search-based Test Case Generation (TCG) for web applications suffers from unstable performance and suboptimal test suite problems due to diversity loss. However, previous diversity metrics mainly only focus on client-side models or server-side code, which are prone to low robustness and poor generalization in practical applications. We propose a diversity-driven TCG method DTester, which can maximize behavior exploration and minimize the test suite size while covering more server-side vulnerable paths. Three diversity metrics (i.e. phenotypic coupling, intent coupling and competitiveness) are proposed to measure the underlying relationship between test cases from user behavior, code logic and test execution history. Moreover, a 3-dimensional weight graph is designed to model association among metrics, which provides fine-grained guidance for the genetic algorithm to generate diverse test cases from the client-side behavior model. Our empirical evaluation on five web applications shows that DTester can efficiently and robustly generate better test suites than the state-of-the-art TCG method. The maximum improvement is $85\%$, $60\%$, $716\%$ and $88\%$ in efficiency, test suite size, diversity and robustness.

This study will take the EVA-BSC evaluation model as the research object, combine Web technology and data analysis and processing technology, choose Spring, SpringMVC and MyBatis as the system development framework under J2EE specification, and use Java language to complete the construction of enterprise performance evaluation system. The system fully considers the actual needs of enterprises, takes BSC as the main body of performance evaluation and EVA as an important evaluation index, improves the performance evaluation mode that only evaluates a single financial index, and takes Web application as the presentation form. Through the convenient and efficient operation, it can realize the effective mining of enterprise value drivers, quickly complete the establishment of enterprise performance evaluation system, and put forward a comprehensive application solution for modern enterprises to scientifically evaluate their strategic positioning.

This paper introduces a full-scale lesson management system called WebLec that is currently used by several universities. Although there are numerous similar systems, WebLec includes multiple features. WebLec has been continuously updated and implemented over three years by the educational staff of a university. In order to raise productivity, a new database processing technique called an Object-based Relational Database was integrated into the system. WebLec has been used since 2002, and in the first half of 2004 the system boasted 3,400 student registrants at Kanagawa University and Shizuoka University, among others. The development techniques and usage in several universities is discussed in detail.

Web-based applications are gaining popularity as they require less client-side resources and are easier to deliver and maintain. However, including the Internet as an integral part of the application also poses new privacy challenges. In particular, recent research revealed that many high profile Web applications might cause sensitive user inputs to be leaked from encrypted traffic due to side-channel attacks exploiting unique patterns in packet sizes and timing. Naive solutions, such as random padding and packet-size rounding, may incur prohibitive communication overhead while still failing to guarantee sufficient privacy protection. In this chapter, we first observe an interesting similarity between this privacy-preserving traffic padding (PPTP) issue and another well-studied problem, privacy-preserving data publishing (PPDP). Based on such a similarity, we present a formal PPTP model encompassing the privacy requirements, padding costs, and padding methods. We then formulate PPTP problems and design efficient heuristic algorithms. Finally, we confirm the effectiveness and efficiency of our algorithms through experiments using real-world Web applications.

A major contributor to the scientific reproducibility crisis has been that the results from homogeneous, single-center studies do not generalize to heterogeneous, real world populations. Multi-cohort gene expression analysis has helped to increase reproducibility by aggregating data from diverse populations into a single analysis. To make the multi-cohort analysis process more feasible, we have assembled an analysis pipeline which implements rigorously studied meta-analysis best practices. We have compiled and made publicly available the results of our own multi-cohort gene expression analysis of 103 diseases, spanning 615 studies and 36,915 samples, through a novel and interactive web application. As a result, we have made both the process of and the results from multi-cohort gene expression analysis more approachable for non-technical users.

As a reliable and portable object-oriented programming language, Java has been used in a variety of practical systems including enterprise servers, smart phones, and embedded systems. Correspondingly, numerous educational institutes have offered Java programming courses to foster Java engineers. In order to assist Java programming educations, we have proposed a Web-based Java Programming Learning Assistant System (JPLAS). JPLAS provides four types of problems to cover the different learning levels of students: 1) element fill-in-blank problem, 2) value trace problem, 3) statement fill-in-blank problem, and 4) code writing problem. To encourage continuous self-studies of the students using JPLAS, the answers will be automatically marked in the server by comparing with the stored correct answers in 1) and 2), and by testing them on JUnit that is a single test framework in 3) and 4), where we have adopted the test-driven development (TDD) method. In this chapter, we introduce the outline of JPLAS and its application results to the Java programming course in Okayama University.