Task-Based and Aggregated Search
Task-based search aims to understand the user's current task and desired outcomes, and how this may provide useful context for the Information Retrieval (IR) process. An example of task-based search is situations where additional user information on e.g. the purpose of the search or what the user already knows about the topic can provide valuable additional evidence that can significantly improve retrieval performance.
Task-based search may be especially useful in cases of aggregated search, also known as integrated search in the digital libraries domain. Aggregated search describes the increasingly common IR paradigm of presenting to the user one result list with information from different document and media types, such as Wikipedia entries, Webpages, user-authored content, images, locations, etc. Research into aggregated search addresses the challenge of when and how to fuse different document and media types, and how to present results to the user. An example of aggregated search is the retrieval of scientific content, which involves searching among different domain-dependent document types and structures (e.g. full articles, short abstracts, tables of content).
This workshop aims to stimulate exploratory research in task-based and aggregated search, and to investigate synergies between these two areas. Submissions are welcome both on each of these areas alone, and on synergies between the two areas. See the call for papers for more information.
The TBAS2012 workshop proceedings are now available for download.
Accepted papers/posters and schedule (click titles for presentations):
- 14h30 - 14h35: Welcome and program, Christina Lioma
- 14h35 - 15h15: Keynote talk: Information Access and Integration in Task-based Heterogeneous Environments, by Prof. Kalervo Järvelin (University of Tampere, Finland).
- 15h15 - 15h40: An Exploration of Retrieval-Enhancing Methods for Integrated Search in a Digital Library, by Diana Ransgaard Sørensen, Toine Bogers and Birger Larsen (Royal School of Library and Information Science, Denmark)
- 15h40 - 16h05: Contextualization from the Bibliographic Structure, by Muhammad Ali Norozi (Norwegian University of Science and Technology, Norway), Arjen P. de Vries (Centrum Wiskunde & Informatica, Netherlands) and Paavo Arvola (University of Tampere, Finland)
- 16h05 - 16h30: Copy and Paste as an Indicator of Relevance: Towards the Development of Intelligent Recommender Systems for Software Engineers, by Adam Grzywaczewski, Rahat Iqbal, John Halloran, Kashif Iqbal and Anne James (Coventry University, UK)
- 16h30 - 16h45: Coffee break and poster presentations:
- Extending Aggregated Search in a Social Sciences Digital Library, by Frank Sawitzki, Philipp Schaer and Daniel Hienert (GESIS - Leibniz Institute for the Social Sciences, Germany)
- Zebra: Searching for Rare Diseases. A Case of Task-Based Search in the Medical Domain, by Radu Dragusin (University of Copenhagen, Denmark), Paula Petcu (Findwise ApS, Denmark), Christina Lioma (University of Copenhagen, Denmark) and Ole Winther (Technical University of Denmark)
- 16h45 - 17h10: Relationship between Usefulness Assessments and Perceptions of Work Task Complexity and Search Topic Specificity: An Exploratory Study, by Peter Ingwersen (Royal School of Library and Information Science, Denmark) and Peiling Wang (University of Tennessee, USA)
- 17h10 - 17h35: Simulating Aggregated Interfaces, by Wei Lu, Qi Wang (Wuhan University, China) and Birger Larsen (Royal School of Library and Information Science, Denmark)
- 17h35 - 18h00: Towards the Systematic Design of IR Systems Supporting Complex Search Tasks, by Thomas Beckers and Norbert Fuhr (Universität Duisburg-Essen, Germany)
- 18h00 - 18h20: Interactive discussion, moderated by Peter Ingwersen
- 18h20 - 18h30: Closing address and best paper/poster award, Birger Larsen