Home

Projects

Active Projects:

  • Information Quality in Social Networks (EU FP7 Grant Proposal with UoR, UoW, UTWENTE, IIITB, and Others)
    With the emergence of new Web based interactive and collaborative environments (e.g. Wikis, Blogs, Voting for Publication, etc.) as commonly used by the next World Wide Web generation known as Web 2.0 / 3.0 or Social Web, the vision of a trustworthy global communication and information space is realized and needs to be shaped. A critical aspect, however, for realizing this vision is the assessment and enforcement of the information quality for information assets, that are to be used for informational or navigational needs or as learning objects. These contents can be generated via collaborative efforts in a distributed, social network of contributors (see Wikipedia as a contemporary example).

    To this extent, the proposal aims at establishing a methodological framework for estimating and improving the information quality and trustworthiness of information assets (e.g. learning objects) as they emerge out of socially interacting or contributing networks of user communities. The foreseen methodology relies on two major considerations: (a) user and social network profile (feedback, citations, interactions, subjectivity, etc.), (b) information asset’s profile (size, coherence, cohesion, plurality / versioning, etc., for both contents and user based semantic tags (if any). The methodology also foresees the establishment of a (semantic) metric (multi-dimensional) space, which enables a holistic as well as partial assessment of information assets.

    The methodological framework will be applied to and tested in a number of use cases - covering many, if not all of the relevant aspects:

    • trustworthiness of other users voting/ rating for information quality assessment,
    • combination and synthesis of high quality learning objects,
    • dissemination, retrieval and learning from high quality blogging communities,
    • improving the quality of (semi-automatic) tagging and
    • semantic annotation of knowledge and learning artefacts.

    Phd. No. 1: NN.
    Phd. No. 2: NN.
    Student No. 1: Ali H. Cevik
    Student No. 2: Abdurrahim Birbilen
    Student No. 3: Engin Kayraklioglu

    • Sub-Project 1: Quality Assessment in the Blog Space

      MSc. No. 1:       Orhan Dagli (07/07-01/08)
      Contact:       Dr.-Ing. Markus Schaal, schaal@cs.bilkent.edu.tr
      Publications:
      • Markus Schaal, Güven Fidan, Roland M. Müller, and Orhan Dagli. Quality Assessment in the Blog Space. The Learning Organization (TLO), Special Issue on Web 2.0 practical implications. To appear.
    • Sub-Project 2: Co-Evaluation of Content and User Qualities by Bayesian Reasoning

      MSc. No. 1: Melihcan Türk
      Contact:       Dr.-Ing. Markus Schaal, schaal@cs.bilkent.edu.tr
      Publications:
      • Markus Schaal. A bayesian approach for small information trust updates. In Proceedings of IeCCS 2006, 2006.

Future Projects:

  • Opinion-mining:
    Opinion search and quality assessment of entities (possibly two separate masters projects)
    Sub-Project 1: Opinion search techniques:
    Development of methods and systems to help people find different expressed opinions on entities such as products, persons, events, organizations, and other named entities (e.g., countries, cities, museums, etc.) on the Web. The project involves: Development of IR techniques for search tasks, such as named entity extraction, opinion-based document ranking.
    Sub-Project 2: Quality assessment of entities:
    Related to the above project the here task is to either aggregate and/or summarise different opinions expressed on entities to reach an a single-point of measure of quality on the named entity, or develop information visualisation techniques to represent multiple viewpoints on the entity of interest.

    Contact: Dr. Murat Karamüftüoğlu, hmk@cs.bilkent.edu.tr

  • Research Aid for Multiple Views on the Related Literature Network

    Contact: Dr. David Davenport, david@cs.bilkent.edu.tr

  • Content Evaluation in Social Networks with Bayesian Belief Updates

    Contact: Dr.-Ing. Markus Schaal, schaal@cs.bilkent.edu.tr

  • Multi-Lingual Collaborative Research Proposal Generator

    Contact: Dr.-Ing. Markus Schaal, schaal@cs.bilkent.edu.tr

  • Distinguish Forgotten Papers from One-Day-Flies

    Contact: Dr. David Davenport, david@cs.bilkent.edu.tr 

  • Learning and Teaching in a Web of Wisdom

    Contact: Dr.-Ing. Markus Schaal, schaal@cs.bilkent.edu.tr

Earlier Projects:

  • User Property Estimation based on Feedback Data Mining
    In this project, implicit or explicit user feedback from e.g. Wikipedia or youtube is mined for user property estimation. We develop a taxonomy for user properties, influence factors for feedback quality, existing and novel models for deviation types and their detection. An existing (Adali et al.) statistical programming environment is currently adapted for the simulation and detection of user properties and deviation types based on rank
    aggregators efficiency. 

    Summer Intern No. 1: Anand Raju
    Summer Intern No. 2: Emre Varol
    Contact: Dr.-Ing. Markus Schaal, schaal@cs.bilkent.edu.tr
    Publications:

    • Xiaohui Lu, Markus Schaal, Sibel Adali, and Anand Kishore Raju. Detecting User Types in Object Ranking
      Decisions. In MEDES '09: Proceedings of the International Conference on Management of Emergent Digital EcoSystems, pages 149-156, New York, NY, USA, 2009. ACM.
  • OLPC - One Laptop Per Child
    Having presented the XO-Laptop to Bilkent in April 2008, both the Department of Computer and Instructional Technology Teacher Education (CTE) and the Collaborative Systems Group showed interest in the conduction of research aligned with OLPC-projects in Turkey. Currently, we are planning to develop a collaborative envirenment for children to
    support educational goals, school organization, and collaborative research.


    MSc. No. 1:     NN.
    Contact: Dr.-Ing. Markus Schaal, schaal@cs.bilkent.edu.tr

  • Collaborative Engineering of Transport Information
    The aim of this project is the community-based semantic aggregation of transport information. Based on a flexible location and transportation model, ontological knowledge is entered, updated and queried by its user community. Both named entities and geo-referencing (based on google map) are considered as basic techniques for collaborative knowledge engineering in this application domain. The location model supports any kind of location and path data to be entered by various users of the system and to be aggregated. Data collection is supported by innovative incentives for users to enter data and implementation of a user friendly system for data entry.
    Current Status of The Project: Architecture and design of the overall system and the location model is completed. We are working on path finding algorithm. We are designing a specialized A* algorithm, in which a heuristic based priority formulation is used. Basic collaborative data extraction methods and Google Maps integration are implemented.

    MSc. No. 1: Kerem Ali Uluğ (07/05 - 06/08)
    MSc. No. 2: NN.
    Contact: Dr.-Ing. Markus Schaal, schaal@cs.bilkent.edu.tr