Chapter 1 Introduction
2.7 AH Research Direction
AH researchers have expanded their research boundary from domain specific applications, employing particular AH techniques, to relate to issues such as shareability and reusability, and the Semantic Web. This section gives a brief overview of the research direction AH researchers have taken over the past few years.
2.7.1 Adaptive Web Systems
From AH systems, the term Adaptive Web has emerged (Brusilovsky and Maybury, 2002). In this paper, they noted that, like pioneering AH research that embraced pre-Web systems, the adaptive Web research has combined a number of different research approaches such as hypertext, user modelling, machine learning, information retrieval, and so on, to develop Web systems that are capable of adapting their behaviour in accordance with the background, goal or interest of individual users or groups of users. In this regard, they pointed out that the main difference was that the scope of AH has been widened to incorporate the technologies such as adaptive content selection and adaptive recommendation (the ability of the system to choose and rank most relevant items, and make recommendations based on individuals or groups of users with similar interests, respectively), as well as mobile generation (adaptation based on an expansion of the user model to respond to the context of a user’s work such as location, time, computer platform and bandwidth (Brusilovsky and Maybury, 2002; Cheverst et al., 2002).
2.7.2 Ontologies in AH
An ontology is defined as “a specification of a conceptualisation” or “a specification of a representational vocabulary for a shared domain knowledge” (Gruber, 1993a). Ontologies originated in the field of Artificial Intelligence, particularly knowledge engineering, to facilitate the shareability and reusability of knowledge (Gruber, 1993b). An ontology can be used to represent the relationships or semantics of data to support the information retrieval process. Ontologies and their technologies are research issues in their own right; therefore, this section and Section 2.7.3 give a concise overview of the use of ontology and its relation to AH research.
The use of ontology in AH is to provide a model for the shared conceptual representation of a domain concept. Together with Web Services and Semantic Web technology, the ontology provides an essential part of the system that makes possible the reusability of the data structure and its content between different components or applications.
2.7.3 Web Services and Semantic Web
Web services are software systems designed to support interoperable machine-to- machine interaction over a distributed network. The Semantic Web research has its primary objective as defining and organising data and its associated relationships in a way its meaning can be understood by software processes rather than people (Berners- Lee et al., 2001). The two technologies are therefore commonly counterparts. While the Semantic Web represents semantically-defined data structures which can be interpreted by the machine, the Web services provide standard means to enable the interoperability between various applications that operate on heterogeneous resources or frameworks (Maneewattana et al., 2005). Examples of Web Services and Semantic Web-based AH Systems are summarised below.
De Bra et al. (2004) proposed a new, modular AH architecture which allows the collaboration between different applications in the creation and maintenance of a user model. These different components communicate with each other via service invocations. Ontologies define the unifying system’s terminology and properties of each system service, and promote the shareability and interoperability among the services. Similarly, Kuruc (2005) suggested sharing a user model between AH applications via the use of Web Service technology. These AH systems have their own domain and adaptation model and make use of User Model Web Service (UMWS) to manipulate their user model.
Aroyo et al. (2004) presented a service-oriented framework for adaptive Web- based systems based on the provision of richer semantics for the adaptive support, the standardisation of user profiling to facilitate adaptation, and the application of reasoning services within distributed Web applications.
In addition, Henze (2005) offered a modular framework for the development and maintenance of the personalised functionalities on the Semantic Web. The Personal Reader framework is a service-based architecture which provides the user interface,
mediates between user requests and available personalisation services, and delivers additional personal recommendations on the viewing context.
Furthermore, Maneewatthana et al. (2005) presented a system called Adaptive Personal Information Environment (a-PIE), a service-oriented framework using Open Hypermedia and Semantic Web technologies to support the shareability and reusability of knowledge in response to the requirements of the users. a-PIE offers users the functionality to search an information space and add and/or manipulate required data into their personal information space without any changes in original information structures.
2.7.4 Remarks on AH Research Direction
AH research has recently been very much about the authoring of adaptable and adaptive hypermedia, applying service-oriented architectures to adaptive Web-based systems, recommender systems and intelligent user interfaces, and the application of the Semantic Web Technologies for adaptive hypermedia and adaptive educational hypermedia. Since 2003, there have been no major steps in establishing new methods for the adaptation of contents and links. Adaptive Web-based systems are no longer single or domain specific applications, but instead modular distributed applications, where new technologies have been implemented so that user models and adaptation rules can be shared and reused amongst multiple distributed applications (De Bra et al., 2004).
Semantic Web technologies can support AH by enhancing existing adaptive techniques and providing an alternative view for adaptation. Ontologies can be applied to describe the system’s terminology and properties of the system for sharing and interoperability among the service-oriented systems. Ontologies can enhance the adaptation of contents by providing richer formal descriptions of content authoring and facilitating the sharing of meanings and semantics of information (or knowledge) between different modular systems (De Bra et al., 2004). As for the AH’s adaptive navigational support technique, ontologies can play a part in the links construction. Additional links can be dynamically inserted based on the open hypermedia concept, which will be described in Chapter 3, in relation to the chosen or suggested ontologies. Ontologies can be exploited to decide links, which would allow words in the hyperdocument to be linked based on the relationships between concepts in the
ontologies. A semantically-derived AH system can suggest ontologies for navigation based on the user profile and the page content the user is navigating. This would enable the navigational links to be adapted based on different users or groups of users.