Abstract
The field of Spatial Humanities has advanced substantially in the past years. The identification and extraction of toponyms and spatial information mentioned in historical text collections has allowed its use in innovative ways, making possible the application of spatial analysis and the mapping of these places with Geographic Information Systems. For instance, automated place name identification is nowadays possible with Named Entity Recognition (NER) systems. Statistical NER methods based on supervised learning, in particular, are highly successful with modern datasets. However, there are still major challenges to address when dealing with historical corpora. These challenges include language changes over time, spelling variations, transliterations, OCR errors, and sources written in multiple languages among others. In this article, considering a task of place name recognition over two collections of historical correspondence, we report an evaluation of five NER systems and an approach that combines these through a voting system. We found that although individual performance of each NER system was corpus dependent, the ensemble combination was able to achieve consistent measures of precision and recall, outperforming the individual NER systems. Additionally, the results showed that these NER system are not strongly dependent on pre-processing and translation to modern English.
Highlights
The exploration of place in texts within the field of Spatial Humanities has advanced substantially in the past years
The Spatial Humanities has advanced very quickly in terms of the analysis of place names mentioned in corpora, as the field grows and scholars look for expedite techniques to support the exploration of large digital datasets, more research is needed to solve particular problems related to the correct identification of place names in historical datasets
While one pathway might be to look for specialized techniques that perform the identification and resolution of place names in historical documents, another possibility is to refine and fine tune Named Entity Recognition (NER) systems/models that are already available
Summary
The exploration of place in texts within the field of Spatial Humanities has advanced substantially in the past years. The identification of geographical places written in historical documents is not carried out in a uniform fashion across the Spatial Humanities and, as said before, it usually relies on manual annotation or the use of one customized NER tool and one main gazetteer The reasons for this are varied, and they might be related to the fact that the Spatial Humanities is still a young field of research that remains highly experimental, and that the challenges involved in creating fully automated processes are not easy to tackle (Gregory et al, 2015; Purves and Derungs, 2015; Wing, 2015). A necessary step to move toward a possible standardization in the methodologies we employ in the Spatial Humanities, to automatically resolve place names in corpora, is the assessment of the performance of different NER systems when used in historical sources
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have