Cysill Ar-lein: A Corpus of Written Contemporary Welsh Compiled from an On-line Spelling and Grammar Checker

15263

Authors

nan

Abstract

AbstractThis paper describes the use of a free, on-line language spelling and grammar checking aid as a vehicle for the collection of a significant (31 million words and rising) corpus of text for academic research in the context of less resourced languages where such data in sufficient quantities are often unavailable. It describes two versions of the corpus: the texts as submitted, prior to the correction process, and the texts following the user’s incorporation of any suggested changes. An overview of the corpus’ contents is given and an analysis of use including usage statistics is also provided. Issues surrounding privacy and the anonymization of data are explored as is the data’s potential use for linguistic analysis, lexical research and language modelling. The method used for gathering this corpus is believed to be unique, and is a valuable addition to corpus studies in a minority language.

Keywords

nan

Full Content

Cysill Ar-lein: A Corpus of Written Contemporary Welsh Compiled from an On-line Spelling and Grammar Checker Delyth Prys, Gruffudd Prys, Dewi Bryn Jones Language Technologies Unit, Bangor University, Bangor, Wales, UK E-mail: {d.prys, g.prys, d.b.jones}@bangor.ac.uk Abstract This paper describes the use of a free, on-line language spelling and grammar checking aid as a vehicle for the collection of a significant (31 million words and rising) corpus of text for academic research in the context of less resourced languages where such data in sufficient quantities are often unavailable. It describes two versions of the corpus: the texts as submitted, prior to the correction process, and the texts following the user’s incorporation of any suggested changes. An overview of the corpus’ contents is given and an analysis of use including usage statistics is also provided. Issues surrounding privacy and the anonymization of data are explored as is the data’s potential use for linguistic analysis, lexical research and language modelling. The method used for gathering this corpus is believed to be unique, and is a valuable addition to corpus studies in a minority language. Keywords: corpus, proofing tools, Welsh, minority language 1. Background The Cysill Ar-lein Corpus is a written corpus of contemporary Welsh created through the novel means of collecting texts from an online spelling and grammar checking service, Cysill Ar-lein. Launched in 2009, Cysill Ar-lein is an online implementation of the popular Cysill Welsh-language spelling and grammar checking tool for Microsoft Windows, first produced in the early 1990s (Hicks, 2004) and updated, repackaged with a compendium of electronic dictionaries, and released as Cysgliad in 2004. Independent studies have since highlighted the popularity of this proofing tool and its role in supporting use of Welsh in computing, online and multimodal environments (Prys, 2008). In 2015, the Cysill Ar-lein interface was made available to external developers to embed in their own websites using an API key available from the Welsh National Language Technologies Portal (2015). Figure 1: The Cysill Ar-lein interface Whilst the original Windows version of Cysill remains part of the commercial Cysgliad package of Welsh-language tools and resources, Cysill Ar-lein was released as a free- to-use service. This was done primarily to enable the service’s developers to gather as large as possible a corpus of contemporary written Welsh from user submissions. A secondary consideration was to ensure that a version of the software was available to all, regardless of their budget or operating system of choice, whilst promoting rather than undermining the sales of the commercial package which funds the continued support and development of the software. For these reasons, unlike the Windows version, the online Cysill Ar-lein service is limited to processing approximately 3000 characters at a time, and its users must accept terms of use that explicitly allow any submitted texts to be used for research purposes. All versions of Cysill do however make use of the same lexicon, part of speech tagger and grammar rules. 2. The Cysill Spelling and Grammar Checking Process Cysill’s spelling and grammar checking …

Content truncated to 500 words

Labeled Text

AbstractThis paper describes the use of a free, on-line language spelling and grammar checking aid as a vehicle for the collection of a significant (31 million words and rising) corpus of text for academic research in the context of less resourced languages where such data in sufficient quantities are often unavailable. It describes two versions of the corpus: the texts as submitted, prior to the correction process, and the texts following the user’s incorporation of any suggested changes. An overview of the corpus’ contents is given and an analysis of use including usage statistics is also provided. Issues surrounding privacy and the anonymization of data are explored as is the data’s potential use for linguistic analysis, lexical research and language modelling. The method used for gathering this corpus is believed to be unique, and is a valuable addition to corpus studies in a minority language.

Publication Details

Paper ID:
15263
Year:
2016
Publication:
TIPSTER
Publication Type:
conf
DOI:
nan