The potential benefit of focused retrieval in relevant-in-context task

Authors:
Paavo Arvola;Johanna Vainio
Affiliations:
University of Tampere, School of Information Sciences, Finland;University of Tampere, School of Information Sciences, Finland
Venue:
INEX'10 Proceedings of the 9th international conference on Initiative for the evaluation of XML retrieval: comparative evaluation of focused retrieval
Year:
2010

Citing 8
Cited 0

Information retrieval and structured documents

Lectures on information retrieval
Cumulated gain-based evaluation of IR techniques

ACM Transactions on Information Systems (TOIS)
Using graded relevance assessments in IR evaluation

Journal of the American Society for Information Science and Technology
Providing consistent and exhaustive relevance assessments for XML retrieval evaluation

Proceedings of the thirteenth ACM international conference on Information and knowledge management
Evaluating relevant in context: document retrieval with a twist

SIGIR '07 Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval
INEX 2007 Evaluation Measures

Focused Access to XML Documents
Focused access to sparsely and densely relevant documents

Proceedings of the 33rd international ACM SIGIR conference on Research and development in information retrieval
Expected reading effort in focused retrieval evaluation

Information Retrieval

Quantified Score

Hi-index	0.00

Visualization

Abstract

This study addresses the Relevant-in-Context retrieval task and seeks justification for systems providing focused answers to be successful in it. Obviously, under some circumstances the full document retrieval is sufficient in finding relevant material effectively. Namely, the Relevant-in-Context retrieval does not bring any improvements in case the retrieved documents are thoroughly (i.e. densely) relevant, or the relevant material is located in the document start. By using the INEX data, we perform a topic-wise analysis focusing on these qualities of the retrieved relevant documents. In addition, we evaluate the submitted INEX runs with various measures, in order to study how different T2I values affect the mutual rankings and measure the systems in locating the relevant material within a document.