2003-08-29 User-controlled link adaptation http://www.ht03.org/papers/pdfs/25.pdf Abstract from http://www.ht03.org/papers/ "This paper introduces an adaptable hypermedia approach applied to adaptive link annotation techniques. The combination of such direct manipulation with automated link annotation affords greater user control over page adaptation. In turn, this direct control better supports user focus in information discovery tasks. Unlike adaptive-only systems, our approach lets users both define multiple topics of interest and then manipulate how these topics` associated links are presented in a page. We discuss how the approach can be applied both to pages viewed as well as to the user's history list, thereby relieving users from the task of either adding to or organizing bookmarks. We describe the prototype developed to support these manipulations, as well as the adaptive architecture developed to support these controls." general problems of hypertext: (there are references for all of these) * disorientation - even in small hypertexts no orientation clues are provided * cognitive overhead - additional effort to maintain several trails/tasks - limited capacity of human short memory * empirical results solutions: * good page design - but this might not match user's mental model * search engines - return pages lacking their surround context [not just on the site, but isolated in concept-space too] * glosses and fluid links - user has to hover over links * bookmarks and history lists - hard to maintain or access Adaptive Hypermedia * provides automated assistance on a user model (annotating, hiding, organising, recommending) * reduce the gap between info and user needs * problems: - users do not have direct control over adaptation - users may not be able to comprehend the mechanism - adaptation may be based on assumptions that are not true The approach: adaptive + adaptable * allow users to directly control interests and adaptation (interface manipulators) * help users identify interesting links * continuous visual feedback * filters of history lists (to help organise) System overview: * there's a server-side thing which adapts the page * the client interface is in flash, this controls the adaptations [insert various algorithms for how to interpret the sliders into a 'relevance' value, see the paper for more.] meaning of single slider on adapation controller: . degree to which a topic belongs to the current user interests meaning of multiple sliders: . either AND or OR requirements: . distinction between relevant, irrelevant links . multiple degrees of relevance and after adapting this, you can show the relevance of a link by altering the link colour, font size. [the demo shows a bunch of sliders in the panel being moved, and a webpage changing link colours as links become more or less relevant. this is an interesting way of establishing a context in a space, exploring the entire page/area at once.] [one of the people on this paper is also responsible for mspace, the prototype of which lets the user explore/establish_content within the space of classical music -- without having to understand the axes. very interesting.] use the same principle for history lists & bookmarks: history . no connection with user interests . hard to access . rarely used bookmarks: . personalised, reflects user interests . hard to maintain (37% don't use them at all) so have a halfway house between the two of these... organise by access time or other info, by topic use the sliders to filter visited nodes --> there's a lot of related work on the subject of bookmarks [but i didn't manage to write it down. have to check the references in the paper and hope it's there.]