A contextual framework for website adaptation
2017-02-16T05:28:39Z (GMT) by
The proliferation of websites and degree of flexibility and freedom allowed when developing websites has resulted in websites with varying navigability. Different approaches have been proposed to optimize websites or to perform website restructuring, mainly in the areas of web mining and web personalization. Existing approaches tend to focus on different methods on improving a website or user navigation without giving due regard to the user and web developer or the characteristics of the website itself. The importance of the website structure in facilitating navigation of a user towards fulfilling her information need is also not given due recognition. In this thesis, we propose a contextual framework for website adaptation that provides a coherent environment for experiential knowledge discovery to improve a user's navigation for satisfYing an information need. The framework unifies the web developer and user and incorporates different interactions with a website. Involving the use of two distinct models that represents each participant, three types of adaptation are proposed in the framework. Each type of adaptation is distinct in its use of the model(s) of information source as well as target audience and resulting output. The first type of adaptation, syntactic adaptation, offers a coarse level of granularity for the adaptation that is performed. The next type, community adaptation, involves both models of information sources. Finally, user adaptation is adaptation at a finer level of granularity that utilizes the resulting output from community adaptation to aid user navigation. A key feature of the framework is that it offers flexibility and extensibility through use of the distinct models representing each participant as well as the types of adaptation. Additionally, the framework is generalized to different categories of websites. Alongside this framework, two categories of evaluation metrics are proposed. The evaluation metrics provides measures for different properties of a website and the resulting output of the adaptation. Collectively, the metrics provide the web developer with the opportunity to analyse the results of the adaptation so that she can make an informed decision on improvements to be made to the website. Using two case studies of different websites, a demonstration of the application of the contextual framework and the evaluation metrics is shown. An analysis of the results shows how the different types of adaptation can improve user navigation as well as aid a web developer to perform website restructuring.