- beyond Recall and Precision
in conjunction with APWeb-WAIM 2009 | April 2-4, 2009 | Suzhou, China

The ENQOIR workshop has been merged with the DBIR workshop. For further news and update refer to the DBIR-ENQOIR website @ http://deke.ruc.edu.cn/dbir09/dbir09.html.
We will have a special session with the accepted papers on the issues of evaluation of ontology-driven information retrieval at the merged workshop.

Call for Papers:

Download CfP as simple text file TXT or as a poster A4:PDF, Three-fold booklet:PDF

In the recent years, a significant research effort has been devoted to ontology-based information retrieval (ObIR). The progress and results in this area offer a promising prospect to improve performance of current information retrieval (IR) systems. Furthermore, existing sparse evaluations of the ObIR tools report improvement compared to traditional IR systems. However, the results lack indications whether this improvement is optimal, causing difficulties to benchmark different ObIR systems. Yet, majority of IR evaluation methods is mainly based on relevance of retrieved information. While additional sophistication of the ObIR tools adds complexity on user interaction to reach improved results. Therefore, standard IR metrics as recall and precision do not suffice alone to measure user satisfaction because of complexity and efforts needed to use the ObIR systems. We need to investigate what ontology properties can even further enhance IR, to assess whether this improvement comes at a cost of interaction simplicity and user satisfaction, etc.

Furthermore, evaluation methods based on recall and precision do not indicate the causes for variation in different retrieval results. There are many other factors that influence the performance of ontology-based information retrieval, such as query quality, ontology quality, complexity of user interaction, difficulty of a searching topic with respect to retrieval, indexing, searching, and ranking methods. The detail analysis on how these factors and their interactions affect a retrieval process can help to dramatically improve retrieval methods or processes.

From other hand, ontology’s ability to capture the content of the universe of discourse at the appropriate level of granularity and precision and offer the application understandable correct information is important. An important body of work already exists in ontology quality assessment field. However, most of ontology evaluation methods are generic quality evaluation frameworks, which do not take into account application of ontology. Therefore there is a need for task- and scenario-based quality assessment methods that, in this particular case, would target and optimize ontology quality for use in information retrieval systems.

In order to promote more efficient and effective ontology usage in IR, there is a need to contemplate on analysis of ontology quality- and value-added aspects for this domain, summarize use cases and identify best practices. Several issues have been put forward by the current research, like the workload for annotation, the scalability, and the balance between the express power and reasoning capability. An approach to holistic evaluation should assess both technological and economical performance viewpoints. An aspect of value creation by semantics-based systems is important to demonstrate that the benefits of the new technology will overwhelm the payout.

The purpose of this workshop is to bring together researchers, developers, and practitioners to discuss experiences and lessons learned, identify problems solved and caused, synergize different views, analyse interplay between ontology quality and IR performance, and brainstorm future research/development directions. Particularly, we strongly encourage submissions dealing with ontology quality aspects and their impact on IR results, evaluation of usability of the ObIR systems, analysis of user behaviour, new evaluation methods enabling thorough and fine-grained analysis of ObIR technological and financial performance, etc.

We invite submissions of two types: regular papers, and research in progress papers. Papers are restricted to a maximum length of 12 pages (including figures, references and appendices). Submissions must conform to Springer's LNCS format. All accepted papers will be published as post-proceedings in a combined APWeb-WAIM'09 workshops volume of Lecture Notes in Computer Science series by Springer. Please submit a paper in the PDF format.