Indri at TREC 2007: Million Query (1MQ) Track

9 years 3 months ago
Indri at TREC 2007: Million Query (1MQ) Track
This work details the experiments carried out using the Indri search engine for the ad hoc retrieval task in the TREC 2007 Million Query Track. We investigate using proximity features for this task, and also explore whether using a simple spelling checker Aspell to correct plausible spelling errors in the noisy queries could help retrieval. Results evaluated by three different approaches are presented. The strength and weakness of introducing Aspell for IR are discussed.
Xing Yi, James Allan
Added 07 Nov 2010
Updated 07 Nov 2010
Type Conference
Year 2007
Where TREC
Authors Xing Yi, James Allan
Comments (0)