Some highlights from the oral paper presentations:
- Le Li and Mark Smucker – Tolerance of Effectiveness Measures to Crowdsourced Relevance Judging Errors. The authors investigate the robustness of established retrieval quality metrics to assessor disagreement. Among the compared metrics, MAP appears to be most resistant to disagreement among judges.
- Matteo Catena et al. – On Inverted Index Compression for Search Engine Efficiency. (Best Paper) The authors demonstrate drastic reductions in average query response time by using modern integer compression schemes on the search engine’s inverted index.
- Matteo Brucato and Danilo Montesi – Metric Spaces for Temporal Information Retrieval. For the class of temporal queries, the authors investigate the temporal scope of queries and documents in order to better estimate document relevance.