An ontology-based fuzzy matching approach to semantic retrieval of historical place names

Abstract

In China Academic Digital Associative Library, there are massive resources recording the history of China. For convenient accessing, services are provided with the feature of information retrieval by geographical dimension, e.g., given a place name under some dynasty, all the information related to that place throughout Chinese history is retrieved. However, the name, boundaries, and administrative level of a place may have changed several times across history. So when the service is asked for information related to some place, all the place names ever used for or related to that place should be identified first. In this paper, we build an ontology model on the geographic information extracted from historical maps, and propose a fuzzy matching algorithm to properly identify places across dynasties. We also present a detailed case study, as well as the experimental results measured by pre-defined metrics, which illustrates the effectiveness of the proposed approach. © 2012 Springer-Verlag.