Logo des Repositoriums

Auflistung nach Autor:in "Bakalov, Fedor"

1 - 7 von 7
  • Konferenzbeitrag
    Adaptive Portals: Adapting and Recommending Content and Expertise
    (16th Workshop on Adaptivity and User Modeling in Interactive Systems, 2008) Nauerz, Andreas; Bakalov, Fedor; König-Ries, Birgitta; Welsch, Martin
    Today, Portals provide users with a central point of access to companywide information. Initially they focused on presenting the most valuable and widely used information to users providing them with quick and efficient information access. But the amount of information accessible quickly grew and finding the right information became more and more complex and time consuming. In this paper, we illustrate options for adapting and recommending content based on user- and context models that reflect users’ interests and preferences and on annotations of resources provided by users. We additionally leverage the entire communitys’ interests, preferences and collective intelligence to perform group-based adaptation. We adapt a Portal’s structure (e.g. navigation) and provide recommendations to be able to reach content being of interest easier. We recommend background in- formation, experts and users with similar interests. We finally construct a Portal’s navigation structure entirely based on the communitys’ behavior. Our main concepts have been prototypically embedded within IBM’s WebSphere Portal.
  • Konferenzbeitrag
    Adaptive Treemap Based Navigation Through Web Portals
    (16th Workshop on Adaptivity and User Modeling in Interactive Systems, 2008) Nauerz, Andreas; Schindler, Sirko; Bakalov, Fedor
    Today, Portals most often present the navigation in form of tree-like structures to users. These trees follow a representation of data with a natural hierarchy. Single nodes are either parents or children. This representation style has several drawbacks: for reaching the leaves users have to traverse rather long paths. During the traversion of the tree users get often lost and, finally, the content relations remain unclear. It is not obvious where the ”most” content is located, too. We regard treemapping (a method for displaying tree-structured data using nested polygons) as a promising technology to overcome these draw- backs. We especially focus on adaptive treemaps displaying different treemaps to different users - always trying to satisfy each single users’ individual needs. Our main concepts have been prototypically embedded within IBM’s WebSphere Portal.
  • Konferenzbeitrag
    A hybrid approach to identifying user interests in web portals
    (9th International Conference On Innovative Internet Community Systems I2CS 2018, 2009) Bakalov, Fedor; König-Ries, Birgitta; Nauerz, Andreas; Welsch, Martin
    Web portals pioneered as one of the earliest adopters of personalization techniques to help users dealing with the problem of information overload. Nowadays they are extensively used as a single-point of access to the vast amount of resources available on the Web and in enterprise intranets. A number of researchers have been investigating the possibilities to enable portals to deliver the content in a highly-personalized manner in order to provide users with a quick and efficient access to the subset of resources relevant to their information needs. However, in order to achieve such a personalization effect, the portal needs accurate and up-to-date information about users, especially the information about their interests. In this paper, we describe a hybrid approach to identifying user interests in Web portals. In our approach, the portal is enabled to "learn" the user interests from the content of visited pages. In addition, it is empowered to provide users with an open access interface to their user models to let them explicitly specify their interests and, in case of incorrectly identified interests, outvote the portal.
  • Konferenzbeitrag
    Link Clouds and User-/Community-Driven Dynamic Interlinking of Resources
    (17th Workshop on Adaptivity and User Modeling in Interactive Systems, 2009) Nauerz, Andreas; Welsch, Martin; Bakalov, Fedor; König-Ries, Birgitta
    During the last years we have observed a shift in the way how content is added to web-based systems. Earlier, dedicated authors were responsible for adding content, today entire communities contribute. As a consequence these systems grow quickly and uncoordinated. New ways had to be found to organize and structure content. Tagging has become one of the most popular techniques to allow users (and entire user communities) to perform this structuring autonomously. But, not only because current tagging systems have their flipsides (e.g. synonyms and polysems lead to littered tag spaces making it difficult for users to find relevant content), we argue that tagging is sometimes an abstraction layer not necessarily needed. In many scenarios users just want to interlink content fragments (re- sources) with each other. In this paper we present an approach allowing users, i.e. the community, to collaboratively define relations between arbitrary content fragments. They can interlink any source with any target. We allow for personal interlinking of resources as well as collaborative interlinking. In the latter case we visualize, for each single resource, available interlinks in what we call link clouds, a concept comparable to tag clouds. We finally leverage the knowledge about the interlinks between resources’ for building personal (or community) navigation structures and for performing content recommendations. The concepts presented are being prototypically implemented within IBM’s WebSphere Portal and can be presented in a live demo at the work- shop.
  • Konferenzbeitrag
    New Tagging Paradigms for Enhancing Collaboration in Web 2.0 Communities
    (17th Workshop on Adaptivity and User Modeling in Interactive Systems, 2009) Nauerz, Andreas; Brück, Matthias; Welsch, Martin; Bakalov, Fedor; König-Ries, Birgitta
    In this paper we present new sophisticated tagging paradigms and their influence on users collaboration behavior and the construction of user– and context–models. We present paradigms like alien tagging which allows one user to apply tags for another user, reputation-based tagging which allows users’ expertise to influence tags’ weights, quantitative tagging which allows users to manually manipulate tags’ weights, anti tagging which allows users to specify ”negative tags”, tag voting to solve the tag space littering problem by e.g. allowing users to vote against tags, tag expiry which allows tags to have a lifetime, contextual tagging which allows tags to be associated to certain context profiles, and so forth and describe how these can be used to refine our models and to perform even more valuable adaptations or to issue more valuable. We also allow for mechanisms to follow users’ tagging ”trails” in order to learn from what they are tagging. All these techniques aim to provide the user with more advanced ways, to add, filter, group and view tags. The concepts presented are currently been prototypically implemented within IBMs WebSphere Portal and can be presented in a live demo at the workshop.
  • Konferenzbeitrag
    An overview of current approaches to mashup generation
    (WM2009: 5th conference on professional knowledge management, 2009) Fischer, Thomas; Bakalov, Fedor; Nauerz, Andreas
    Mashups allow users to bring together data and services from various web applications in order to create a new integrated tool that serves their needs. In the last few years, a variety of mashups frameworks has been proposed that promise to simplify the mashup creation process so that every user is able to create mashups. In this paper, we give an overview about these approaches and identify their limitations. The main insight is that the average user will not possess the necessary skills to create mashups that meet his needs with these tools. We therefore propose that a tool is needed that allows for the automatic ad-hoc generation of mashups.
  • Konferenzbeitrag
    Towards an Automatic Service Composition for Generation of User-Sensitive Mashups
    (16th Workshop on Adaptivity and User Modeling in Interactive Systems, 2008) Fischer, Thomas; Bakalov, Fedor; Nauerz, Andreas
    In today’s Web 2.0, mashups allow users to bring together data and services from various Web applications in order to create a new integrated tool that serves their needs. Nowadays, there is an increasing number of frameworks that provide users with a GUI environment to manually assemble different data sources and services into a mashup. However, in order to create such tools, the user must possess a certain level of technical knowledge. In this paper, we introduce a framework that automatically selects and combines Web services to create mashups. We also describe the user model that stores knowledge about user interests and expertise, which are used by the framework in order to generate mashups tailored to the needs of individual users.