Statistical principal components analysis for retrieval experiments

摘要: In this article, the statistical principal components analysis (PCA) is proposed as a method for performance comparisons of different retrieval strategies. It shown that PCA can reveal implicit relations among systems across information needs (i.e., queries, topics). For illustration, TREC 12 robust track data have been reevaluated by and to easily are hard see with traditional techniques. Therefore, promises uniform evaluation framework be used large-scale experiments. addition mean average precision (MAP) measure, relative analytic distance (RAD) new summary measure based on same notion introduced PCA.