Espresso: Leveraging Generic Patterns for Automatically Harvesting Semantic Relations

摘要: In this paper, we present Espresso, a weakly-supervised, general-purpose, and accurate algorithm for harvesting semantic relations. The main contributions are: i) method exploiting generic patterns by filtering incorrect instances using the Web; ii) principled measure of pattern instance reliability enabling algorithm. We an empirical comparison Espresso with various state art systems, on different size genre corpora, extracting general specific Experimental results show that our exploitation substantially increases system recall small effect overall precision.