Trailist - An Open Framework for Supporting Trail-based Information Access

Research project P 14006 Trailist Siegfried REICH 11.10.1999 People are confronted with ever growing amounts of information. This makes it increasingly difficult to navigate vast information spaces in order to find the appropriate information. Basing access to and retrieval of information on user trails would allow people not only to better manage their personal information spaces but also to ask questions such as "where do other people go from here", "what else should I read? " or "how did we come to that conclusion?". Consequently, users could manage their individual information spaces more efficiently. Furthermore, the fact that trails are built with information about the users` browsing paths and activities makes them well suited for collaborative applications where users with similar interests are to be matched. Since the Greek hero Theseus found his way out of the maze by fastening a thread to the entrance and unwind it to find his way out, we are familiar with the concept of user trails as a means for navigating unknown territories. In hypertext, guided tours and trails are part of the repertoire of concepts that can be used to assist users in navigating vast amounts of information. In hypertext, trails are collections of trail marks. Trail marks are steps that users have undertaken in pursuing a certain task, hence these trail marks have a number of valuable properties, these include the associated activity, the duration, or location to name just a few. Trailist proposes to build an open framework that will support the construction of trail-based applications. This proposal addresses the following issues: * Acquisition of trails: this aspect concerns the creation of trail information. It includes issues such as the preservation of privacy, the notion of trails themselves, i.e. what composes a trail, when do trails end, etc. * Processing of user requests: the more accurate the replies to the users` requests needs to be, the more complex the processing will become. Therefore, different paradigms such as software agents offer themselves for application in this domain. The framework will allow communication between agents, asynchronous message passing, etc. Furthermore, the design and implementation of the open framework will form a major part of the proposal, With "open" we refer to the properties of extensibility at runtime, adaptability to various applications and also the number of users and data which should be unlimited. The proposed framework will comprise general components that can be re-used and adapted to specific requirements. The usability of the framework and the concept of user trails will be evaluated by two sample application prototypes: * A trail-enabled Web browser that will allow people to record their browsing history. This will allow retrieval of documents based on the performed activity thereby enabling improved navigation of vast information spaces. Furthermore, software agents will mine the trail data in order to match users with similar interests. * A personal agent that will automatically keep track of a user`s individual tasks such as e-mailing a document to a colleague, printing a document out, etc. In order to demonstrate this facility, a sample set of desktop applications shall be integrated with the framework. The usability of the framework will be measured along the dimensions of usability, distribution, integration, conformance to standards, manageability, performance, reliability and scalability. The concept of user trails will be evaluated by testing the two prototypes with user groups composed of IT professionals. The evaluation will be performed using questionnaires.