Natural Language Processing for Historical Texts

Piotrowski, Michael

doi:10.2200/s00436ed1v01y201207hlt017

Cited by 115 publications

(81 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A common approach to deal with the high amount of variance often found in this type of data is to perform spelling normalization (Piotrowski, 2012), which is the mapping of historical spelling variants to standardized/modernized forms (e.g. vnd → und 'and').…”

Section: Introductionmentioning

confidence: 99%

Learning attention for historical text normalization by learning to pronounce

Bollmann

Bingel

Søgaard

2017

Proceedings of the 55th Annual Meeting of the Association For Computational Linguistics (Volume 1: Long Papers)

View full text Add to dashboard Cite

Automated processing of historical texts often relies on pre-normalization to modern word forms. Training encoder-decoder architectures to solve such problems typically requires a lot of training data, which is not available for the named task. We address this problem by using several novel encoder-decoder architectures, including a multi-task learning (MTL) architecture using a grapheme-to-phoneme dictionary as auxiliary data, pushing the state-of-theart by an absolute 2% increase in performance. We analyze the induced models across 44 different texts from Early New High German. Interestingly, we observe that, as previously conjectured, multi-task learning can learn to focus attention during decoding, in ways remarkably similar to recently proposed attention mechanisms. This, we believe, is an important step toward understanding how MTL works.

show abstract

Section: Introductionmentioning

confidence: 99%

Learning attention for historical text normalization by learning to pronounce

Bollmann

Bingel

Søgaard

2017

Proceedings of the 55th Annual Meeting of the Association For Computational Linguistics (Volume 1: Long Papers)

View full text Add to dashboard Cite

show abstract

“…Minimal transformations imply orthographic normalization and in some cases syntactic translation rules (Piotrowski 2012). Less attention is paid to semantics and the conceptual space (thus implicit knowledge) which changed during the years (Vertan 2010) (Vertan and v. Hahn 2014a).…”

Section: Automatic Processing Of Historical Textsmentioning

confidence: 99%

On the annotation of vague expressions: a case study on Romanian historical texts

Dinu¹,

Hahn²,

Vertan³

2017

Proceedings of the Workshop on Language Technology for Digital Humanities in Central and (South-)Eastern Europe

View full text Add to dashboard Cite

Current approaches in Digital .Humanities tend to ignore a central aspect of any hermeneutic introspection: the intrinsic vagueness of analyzed texts. Especially when dealing with historical documents neglecting vagueness has important implications on the interpretation of the results. In this paper we present current limitation of annotation approaches and describe a current methodology for annotating vagueness for historical Romanian texts.

show abstract

“…However, according to [8], TEI is very complex, if it is compared to other similar element tagging, such as Dublin Core, METS and MARC, but the complexity is inherent in the task of annotating historical document. Reference [8] illustrates the encoding of a primary source using TEI.…”

Section: A Tei P5 Guidelinesmentioning

confidence: 99%

“…Reference [8] illustrates the encoding of a primary source using TEI. The source"s name is Alteres Landbuch.…”

Section: A Tei P5 Guidelinesmentioning

confidence: 99%

See 1 more Smart Citation

Using TEI XML Schema to Encode the Structures of Sarawak Gazette

Tze-Min¹,

Ranaivo-Malançon²

2015

IJSSH

View full text Add to dashboard Cite

Abstract-Automatic extraction of information from old printed documents which have been digitised injudiciously will end up with a lot human corrections. To overcome the problem, one possible solution is to annotate the documents with some markups. This paper presents the encoding of the digitised sample of Sarawak Gazette published from 1903 until 1939 using the standard TEI XML schema. The output of the work is a set of six TEI XML templates that is considered to represent the different layout structures found in the studied samples.Index Terms-Data structure, layout analysis, metadata, TEI P5 schema.

show abstract

Natural Language Processing for Historical Texts

Cited by 115 publications

References 70 publications

Learning attention for historical text normalization by learning to pronounce

Learning attention for historical text normalization by learning to pronounce

On the annotation of vague expressions: a case study on Romanian historical texts

Using TEI XML Schema to Encode the Structures of Sarawak Gazette

Contact Info

Product

Resources

About