The Polish Sejm Corpus / Polski Korpus Sejmowy

The Polish Sejm Corpus (PSC) is a large (300M segments) collection of stenographic transcripts of Polish Sejm sittings from 1-6 terms of office.

The first edition of the PSC was prepared in October 2011 and contains transcripts of Sejm sessions saved in TEI P5 format. The resource contains automatically created annotation of the following kind:

• utterance-level segmentation,
• tokenization,
• lemmatization,
• disambiguated morphosyntactic description,
• syntactic words,
• syntactic groups,
• named entities.

Corpus data

Whole corpus

• Interpellations and questions terms 3-6, sittings terms 1-6, 25 GB

Divided by term and document type

• Interpellations and questions Term 3, (1997–2001), 1.7 GB

• Interpellations and questions Term 4, (2001–05), 2.4 GB

• Interpellations and questions Term 5, (2005–07), 2.0 GB

• Interpellations and questions Term 6, (2007–11), 4.8 GB

• Sittings Term 1, (1991–93), 0.9 GB

• Sittings Term 2, (1993–97), 2.6 GB

• Sittings Term 3, (1997–2001), 2.8 GB

• Sittings Term 4, (2001–05), 3.4 GB

• Sittings Term 5, (2005–07), 1.4 GB

• Sittings Term 6, (2007–11), 3.1 GB

Publications

Maciej Ogrodniczuk. The Polish Sejm Corpus. In Proceedings of the Eighth International Conference on Language Resources and Evaluation, LREC 2012, pages 2219–2223, Istanbul, Turkey, 2012. European Language Resources Association (ELRA).

Searching the corpus

Online search of the corpus is currently down; please use the Poliqarp image of the corpus for the moment: http://sejm.nlp.ipipan.waw.pl/static/PSC_poliqarp.tar.gz.