In this paper we investigate about automated extraction of author lists in the domain of scientific digital libraries. It is given a list of known “seed” authors and we aim to extract complete lists of co-authors from Web pages in arbitrary format. We adopt a methodology embedding domain knowledge in a unique “meta-wrapper”, not requiring training, with negligible maintenance costs and based on the combination of several extraction techniques. Such methods are applied at the structural level, at the character level and at the annotation level. We describe the methodology, illustrate our tool, compare with known approaches and measure the accuracy of our techniques with proper experiments.
A Domain Meta-wrapper Using Seeds for Intelligent Author List Extraction in the Domain of Scholarly Articles
Cauteruccio, F.;
2013-01-01
Abstract
In this paper we investigate about automated extraction of author lists in the domain of scientific digital libraries. It is given a list of known “seed” authors and we aim to extract complete lists of co-authors from Web pages in arbitrary format. We adopt a methodology embedding domain knowledge in a unique “meta-wrapper”, not requiring training, with negligible maintenance costs and based on the combination of several extraction techniques. Such methods are applied at the structural level, at the character level and at the annotation level. We describe the methodology, illustrate our tool, compare with known approaches and measure the accuracy of our techniques with proper experiments.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
