[Corpora-List] POS-tagger maintenance and improvement

Rayson, Paul rayson at exchange.lancs.ac.uk
Wed Feb 25 22:25:08 CET 2009


Hi Adam,

Presumably you are intending to be provocative there! It is difficult for universities to get research funding to improve POS taggers for English in particular. Why would funding agencies provide resources for this? It is a solved problem isn't it!! :-) The last time UCREL had large funding indirectly for CLAWS was on the BNC enhancement project in 1996: http://ucrel.lancs.ac.uk/projects.html#bnce In that project we also developed the template tagger as a more powerful patching tool for CLAWS, see papers listed at http://ucrel.lancs.ac.uk/claws/

It is certainly feasible for users of CLAWS to add their own supplementary rule files and dictionary files in order to improve the tagging. Email me off list for info.

We are still interested in receiving feedback from users of CLAWS and this has resulted in a number of papers recently as follows. I think the interesting stuff happens when you try to apply standard tools to non-standard language e.g. learner data, historical text, dialect corpora:

Beal, J., Corrigan, K., Smith, N. and Rayson, P. (2007) Writing the Vernacular: Transcribing and Tagging the Newcastle Electronic Corpus of Tyneside English. Studies in Variation, Contacts and Change in English. Volume 1. Research Unit for Variation, Contacts and Change in English (VARIENG), University of Helsinki. http://www.helsinki.fi/varieng/journal/volumes/01/beal_et_al/

Rayson, P., Archer, D., Baron, A., Culpeper, J. and Smith, N. (2007). Tagging the Bard: Evaluating the accuracy of a modern POS tagger on Early Modern English corpora. In proceedings of Corpus Linguistics 2007, July 27-30, University of Birmingham, UK. http://ucrel.lancs.ac.uk/people/paul/publications/RaysonEtAl_CL2007.pdf

We've also got a forthcoming paper at the ICAME pre-conference workshop on 'Errors and disfluencies in spoken corpora' with Joanna Jendryczka-Wierszycka titled "Applying native language trained annotation tools to non-native spoken corpora" relating to applying CLAWS to spoken learner (LINDSEI) material.

For completeness, there are other papers that I refer to in my reply to a previous question of yours about tagging spoken data on the Corpora list:

https://mailman.uib.no/public/corpora/2005-July/001363.html

Regards,

Paul.

Dr. Paul Rayson

Director of UCREL

Computing Department, Infolab21, South Drive, Lancaster University, Lancaster, LA1 4WA, UK.

Web: http://www.comp.lancs.ac.uk/computing/users/paul/ <http://www.comp.lancs.ac.uk/computing/users/paul/>

Tel: +44 1524 510357 Fax: +44 1524 510492

From: corpora-bounces at uib.no [mailto:corpora-bounces at uib.no] On Behalf Of Adam Kilgarriff Sent: 25 February 2009 11:16 To: Corpora List Cc: Sue Atkins; Valerie GRUNDY; Patrick Hanks Subject: [Corpora-List] POS-tagger maintenance and improvement

All,

My lexicography colleagues and I use POS-tagged corpora all the time, every day, and very frequently spot systematic errors. (This is for a range of languages, but particularly English.) We would dearly like to be in a dialogue with the developers of the POS-tagger and/or the relevant language models so the tagger+model could be improved in response to our feedback. (We have been using standard models rather than training our own.) However it seems, for the taggers and language models we use (mainly TreeTagger, also CLAWS) and also for other market leaders, all of which seem to be from Universities, the developers have little motivation for continuing the improvement of their tagger, since incremental improvements do not make for good research papers, so there is nowhere for our feedback to go, nor any real prospect of these taggers/models improving.

Am I too pessimistic? Are there ways of improving language models other than developing bigger and better training corpora - not an exercise we have the resources to invest in? Are there commercial taggers I should be considering (as, in the commercial world, there is motivation for incremental improvements and responding to customer feedback)?

Responses and ideas most welcome

Adam Kilgarriff -- ================================================ Adam Kilgarriff http://www.kilgarriff.co.uk Lexical Computing Ltd http://www.sketchengine.co.uk Lexicography MasterClass Ltd http://www.lexmasterclass.com Universities of Leeds and Sussex adam at lexmasterclass.com ================================================

-------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/html Size: 10976 bytes Desc: not available Url : https://mailman.uib.no/public/corpora/attachments/20090225/346d18da/attachment.txt



More information about the Corpora mailing list