Federico Boschetti

Improving OCR accuracy for classical critical editions

This paper describes a work-flow designed to populate a digital library of ancient Greek critical editions with highly accurate OCR scanned text. While the most recently available OCR engines are now able after suitable training to deal with the polytonic Greek fonts used in 19th and 20th century editions, further improvements can also be achieved with postprocessing. In particular, the progressive multiple alignment method applied to different OCR outputs based on the same images is discussed in this paper.

Citations in the Digital Library of Classics: Extracting Canonical References by Using Conditional Random Fields

Scholars of Classics cite ancient texts by using abridged citations called canonical references. In the scholarly digital library, canonical references create a complex textile of links between ancient and modern sources reflecting the deep hypertextual nature of texts in this field. This paper aims to demonstrate the suitability of Conditional Random Fields (CRF) for extracting this particular kind of reference from unstructured texts in order to enhance the capabilities of navigating and aggregating scholarly electronic resources. In particular, we developed a parser which recognizes word level n-grams of a text as being canonical references by using a CRF model trained with both positive and negative examples.

Restructuring a Taxonomy of Literary Themes and Motifs for More Efficient Querying

In this paper we describe ongoing work in the restructuring of a tagset originally organised as a taxonomy and used to annotate literary themes and motifs in a corpus of classical works of poetry from a number of different traditions. We show how such a tagset can be rendered more efficient and useful through the appropriation of ideas and techniques from lexical semantics and ontology design. The newly redesigned tagset is described with examples showing how the new design is much more expressive than the old taxonomy; furthermore, an example query is described in order to demonstrate how more refined semantic searches can be carried using the new version of the taxonomy. The final result is, we hope, a resource that will be useful not only for the specific project for which it was developed but one that is well-designed and well-documented enough to be of use for other similar semantic annotation tasks. DOI: http://dx.doi.org/10.14195/2182-8830_4-2_1

Computational Linguistics and Language Physiology: Insights from Arabic NLP and Cooperative Editing

Computer processing of written Arabic raises a number of challenges to traditional parsing architectures on many levels of linguistic analysis. In this contribution, we review some of these core issues and the demands they make, to suggest different strategies to successfully tackle them. In the end, we assess these issues in connection with the behaviour of neuro-biologically inspired lexical architectures known as Temporal Self-Organising Maps. We show that, far from being language-specific problems, issues in Arabic processing can shed light on some fundamental characteristics of the human language processor, such as structure-based lexical recoding, concurrent, competitive activation of output candidates and dynamic selection of optimal solutions.