A Poor Man’s Approach to CLEF

Abstract

The primary goal of our participation in CLEF is to acquire experience with supporting cross-lingual retrieval. We submitted runs for all four target languages, but our main interest has been in the bilingual Dutch to English runs. We investigated whether we can obtain a reasonable performance without expensive (but high quality) resources; we have used only ‘off-the-shelf’, freely available tools for stopping, stemming, compound-splitting (only for Dutch) and translation. Although our results are encouraging, we must conclude that a poor man’s approach should not expect to result in rich men’s retrieval results.