this article we show the power of this approach. On the one hand, UWE prescribes how to build navigation and presentation models for Web applications defining therefore special UML stereotyped modeling elements and tagged values. On the other hand, we show how these Web specific navigation and presentation models can be supplemented by other views using the variety of UML diagram types and UML modeling elements. Our current research activities focus on the dynamic aspects of the design of Web applications, such as task modeling and modeling of Web scenarios, graphical representation of the distribution of Web components and semi-automatic generation of Web applications based on design models

We propose a precise UML-based authoring method for Web applications. This authoring method is part of the UML-based Web Engineering (UWE) approach. Starting with a requirement analysis done by use cases technique, it focuses on the design phase. The conceptual model of the application is used as guideline for modeling the navigation space. From the navigation space model we derive, in a next step, a navigation structure model which shows how to navigate through the navigation space using access elements like indexes, guided tours, queries and menus. Finally, a presentation model is constructed based on the navigation structure model. It provides an appropriate UML notation to support sketching and storyboarding. In addition, we suggest to use UML interaction diagrams to represent the presentation flow. During the whole development process we identify steps that can be performed in an automatic way thus providing the basis for a generation mechanism for Web application design. The different models of the design process are represented by using a UML conform extension of UML for Web applications (UML profile). The authoring process is illustrated with an example: a Web--based conference review system. The strength of the presented Web engineering approach is given by the fact that we use exclusively the UML notation and techniques. Moreover, our specification of constraints with OCL (part of UML) allows augmenting the exactitude of the models. In the same way our methodology has a high degree of precision in the description of guidelines of the authoring process of Web application, which can even partially be automated. Keywords: Web Engineering, Unified Modeling Language, Web applications, Authoring Process, Design Method, Systematic Development, UML Extension 1

Abstract. Although adaptive applications are increasing in popularity, there are only a few approaches that focus on their generalization or the specification of a reference model. Trying to fill this gap, this paper presents a reference model for adaptive hypermedia applications, similar to AHAM. The main novelty of our approach is an object-oriented specification written in UML (Unified Modeling Language) which integrates both an intuitive visual representation and a formal unambiguous specification in OCL (Object Constraint Language). Our reference model is defined as an extension of the Dexter Hypertext Reference Model including user modeling aspects and rule-based adaptation mechanisms.

In this paper we present a method for the semiautomatic transition from the design models of a Web application to a running implementation. The design phase consists of constructing a set of UML models such as the conceptual model, the navigation model and the presentation model. We use the UML extension mechanisms, i.e. stereotypes, tagged values and OCL constraints, thereby defining a UML Profile for the Web application domain. We show how these design models can automatically be mapped to XML documents with a structure conforming to their respective XML Schema definitions. Further on we demonstrate techniques how XML documents for the conceptual model are automatically mapped to conceptual DOM objects (Document Object Model). DOM objects corresponding to interactional objects are automatically derived from conceptual DOM objects and/or other interactional DOM objects. The XSLT mechanism serves to transform the logical presentation objects representing the user interface to physical presentation objects, e.g. HTML or WAP pages. Finally we present a production system architecture for Web applications using the XML publishing framework Cocoon which provides a very flexible way to generate documents comprising XSLT and XSP (eXtensible server pages) processors.

Abstract. The rising interest for developing systems in web environments has produced that the research community poses the necessity of offering methodological proposals in order to give a suitable reference when a web system is produced. In the last years, several methodological proposals have appeared: OOHDM (Object Oriented Hypermedia Design Method, UWE (UML Based Web Engineering), OOH (Object-Oriented Hypermedia Method), WSDM (Web Site Design Method) or WebML (Web Modelling Language) are only some examples. However, there is not a standard and world wide accepted methodology. Each methodology proposes its own techniques and models. This fact has produced that some studies have been made in order to compare them. These comparatives studies show that there are some gaps in web engineering. This paper presents these gaps and introduces NDT (Navigational Development Techniques). NDT is an approach which offers a different development process that starts with requirements treatment and allows to get design models using a systematic process. It tries to solve some of these gaps detected in web engineering. In order to introduce how NDT confronts these gaps, the paper presents NDT development process, a global vision of its objectives and its tool case, NDT-Tool. Also, this paper enumerated some real projects developed by NDT and NDT-Tool in real companies.

Adaptive Hypermedia Systems (AHS) affect the way most of the Web-based applications are developed and used. One of the application domains affected by AHS is the domain of Web-based Support Systems (WSS), which is part of the more general online information systems domain. Since the very beginning of the research on AHS, different systems have been developed for different domains and application areas (such as educational systems, online information systems, etc.). Besides some general features, which are expected to be present in any software system, AHSs have their own extra features that make them different from non-adaptive systems. Through a careful study of a large number of adaptive hypermedia systems, a hierarchy of primary features based on which one can evaluate different adaptive web systems is proposed. An evaluation weighting scheme is also proposed for the given features. A comparative analysis is carried out for 4 major adaptive hypermedia systems (AHA!, InterBook, SETA and SeAN). The results of our detailed evaluation put AHA! at the top followed by SETA, SeAN and InterBook, respectively. 1

Abstract. In this paper, we describe the design steps of extending LAOS, a five-layer framework for generic adaptive web learning authoring, by adding a social layer to capture (and adapt) information from 1) collaborative authoring (i.e. editing the content of other learners, describing the content using tags, rating the content, and commenting on the content, etc); and 2) authoring for collaboration (i.e., adding authors ’ activities, such as defining groups of authors, subscribing to other authors, etc). Moreover, the paper presents MOT 2.0, an adaptive E-learning 2.0 system, which is built on the proposed reference model, and finally, we report on our evaluations to validate the new Social Layer by comparing MOT 2.0 with its predecessor, MOT 1.0.

Abstract. Reusable adaptation specifications for adaptive behaviour has come to the forefront of adaptive research recently, with EU projects such as GRAPPLE 1, and PhD research efforts on designing an adaptation language for learning style specification [ 28]. However, this was not the case five years ago, when an adaptation language for adaptive hypermedia (LAG) was first proposed. This paper describes the general lessons learnt during the last five years in designing, implementing and using an adaptation language, as well as the changes that the language has undergone in order to better fulfil its goal of combining a high level of semantics with simplicity, portability as well as being flexible. Besides discussing these changes based on some sample strategies, this paper also presents a novel authoring environment for the programming-savvy adaptation author, that applies feedback accumulated during various evaluation sessions with the previous set of tools, and its first evaluation with programming experts. Keywords: Adaptive Hypermedia, Adaptation Language, LAG, LAOS. 1

Abstract. Nowadays, MDA is gaining popularity as a feasible way to develop software in Web environments. As a consequence, several tools from both academic and industrial contexts, offer their own MDA processes for producing Web Applications. OO-Method is an object-oriented method that produces software systems by means of its MDA implementation, OlivaNOVA. This tool has been broadly tested in industry with real desktop applications. However, it lacks the expressivity needed to accurately describe Web Systems. OOWS is the web-oriented extension of OO-Method with was developed to solve this problem. This work, presents a MDA development environment that combines OO-Method and OOWS. This environment produces fully functional Web Applications that integrate the business logic generated by OlivaNOVA with a Web Interface produced from OOWS models. The followings tools are introduced to support OOWS development process: (1) an Eclipse-based modeller to edit OOWS models visually (2) a Web Interface Framework that is based on Software Factories philosophy in order to reduce the abstraction level between conceptual models and the code to be generated and (3) a set of Model-to-Text transformations that allows the automatic generation of a Web Interface from models. This work also describes a strategy to include the OlivaNOVA development process into the new MDA development environment. 1

Abstract La inclusión de políticas de personalización en el marco de métodos hipermediales influye en todos los niveles de modelado de una aplicación web. Desde OO-H mantenemos que esta política de personalización debe facilitar la definición de aspectos no conocidos a priori, lo que a su vez obliga a estos métodos a posibilitar la adaptación dinámica de las aplicaciones ante cambios imprevistos en el entorno. Este artículo presenta, en el contexto del Método Hipermedial Orientado al Objeto (Object Oriented Hypermedia (OO-H)), un marco de personalización que cumple estas características y que se vertebra en torno a cuatro elementos fundamentales: (1) un conjunto de actividades de diseño que captura la parte estática de la personalización, (2) un mecanismo de especificación de reglas, definidas en base a una plantilla XML, que constituye el modelo de personalización dinámico (3) una arquitectura de ejecución que permite el desacoplamiento de este modelo dinámico respecto al resto de modelos y minimiza el efecto de cambios en la estrategia, y (4) un repositorio extensible que incluye un conjunto de mecanismos de registro de la actividad del usuario en el sistema. La posibilidad de extensión de este repositorio facilita su adaptación a las características particulares de la aplicación modelada. 1 Introducción