National Corpus of Polish

The National Corpus of Polish, known in Polish as Narodowy Korpus Języka Polskiego and commonly abbreviated NKJP, is a national-scale linguistic corpus and online research resource devoted to the Polish language. It brings together texts from multiple institutional collections and makes them searchable for linguistic, lexicographic, educational, and computational purposes. Rather than being a conventional consumer brand, NKJP functions as a public-facing research platform and a shared scholarly infrastructure. The initiative was created by four partners: the Institute of Computer Science of the Polish Academy of Sciences, the Institute of the Polish Language of the Polish Academy of Sciences, Polish Scientific Publishers PWN, and the Department of Computational and Corpus Linguistics at the University of Łódź. The collaboration combined pre-existing corpus-building efforts that had developed separately at the participating institutions. The project was registered as a research-and-development undertaking of Poland’s Ministry of Science and Higher Education, reflecting its national role in documenting and analyzing Polish. Its planned full collection was designed to exceed one billion words. Within that broader resource, a carefully balanced subcorpus of approximately 300 million words was established to provide a more representative picture of Polish usage across genres and periods. A manually annotated corpus of approximately one million words was also released under an open license, supporting more detailed grammatical and computational analysis. The corpus is intended to cover both established and contemporary language, rather than concentrating on a single literary or media genre. The source material includes classic Polish literature, general daily newspapers, specialist periodicals and journals, recorded or transcribed conversations, short-lived publications, and Internet texts. This mixture allows researchers to investigate standard and colloquial usage, historical continuity, specialist terminology, emerging expressions, phraseology, and differences between written and spoken language. Corpus evidence can show how a word or phrase is used in context, what meanings and grammatical functions it takes, and how its frequency or distribution changes over time. NKJP has been associated with several search environments. The Poliqarp interface supports searches for words, phrases, grammatical forms, and more complex sequences, including regular-expression-style queries and morphologically expanded searches. PELCRA provides access to a substantially larger combined collection drawing on several institutional corpora, with downloadable results, visualization features, and tools for examining time series and lexical or phraseological patterns. These interfaces make the underlying data useful not only to specialist corpus linguists but also to lexicographers, language teachers, translators, journalists, students, and developers of language technologies. The corpus’s importance rests on its collaborative and representative design. By joining collections from academic and publishing organizations, it created a common reference point for Polish-language research and reduced dependence on isolated institutional datasets. Its open and online components also broadened access to evidence that would otherwise require specialized software or direct contact with individual archives. The available reference material does not establish a commercial ownership structure, a conventional headquarters, or a corporate parent; NKJP is best understood as an institutional research consortium and digital language resource.

The National Corpus of Polish grew out of several independent corpus projects developed by Polish academic and publishing institutions. The Institute of the Polish Language of the Polish Academy of Sciences created an early corpus that was not initially publicly available. Polish Scientific Publishers PWN later developed its own corpus, while the PELCRA group at the University of Łódź assembled another important collection. The Institute of Computer Science of the Polish Academy of Sciences also developed a separate corpus, with a strong orientation toward computational processing and language technology. These collections were created for related purposes but were initially maintained by separate teams. Their consolidation offered a way to combine complementary strengths: linguistic expertise, publishing data, computational infrastructure, and experience with corpus query systems. In 2006, the four groups joined forces to establish the National Corpus of Polish. The project was registered as a research-and-development project of the Ministry of Science and Higher Education, giving it an explicitly national research mandate rather than the status of a single institution’s local database. The consortium planned a complete resource of more than one billion words. The scale was significant for Polish corpus linguistics because it could support both general frequency research and investigation of comparatively rare terms, constructions, and phraseological patterns. At the same time, the project identified the need for a carefully balanced collection. A balanced subcorpus of roughly 300 million words was prepared to provide a more controlled representation of genres and usage contexts, while a manually annotated corpus of around one million words was made available under an open license for detailed linguistic and computational work. The corpus assembled material from a wide range of sources. Literary works supplied historical and culturally significant examples; daily newspapers contributed current general-language usage; specialist magazines and journals represented technical and professional vocabulary; conversation transcripts captured spoken language; and ephemeral and Internet texts documented less formal, rapidly changing forms of expression. This breadth made the resource useful for studying not only what words mean but also how their meanings, grammatical roles, collocations, and frequencies vary by genre and period. Online access became a central part of the project’s public identity. The Poliqarp search engine enabled users to query words and phrases, retrieve grammatical variants, use regular-expression-like patterns, and examine sequences such as combinations of nouns and adjectives. Its architecture also supported local or offline execution for some operations. PELCRA offered a broader analytical environment drawing on three major corpus collections associated with the project’s partners. It supported spreadsheet downloads, morphological and spelling-related query options, flexible phrase searches, visual exploration of frequency or registry patterns, and time-series generation. Through these developments, NKJP became a shared reference infrastructure for Polish corpus research. Its users include linguists, lexicographers, language educators, translators, computational-language researchers, and others who need authentic examples rather than isolated dictionary definitions. The available material does not document a later corporate restructuring, commercial acquisition, or conventional executive hierarchy. The project is therefore best characterized as an institutional consortium and a digital scholarly platform whose principal historical achievement was the integration and public presentation of major Polish-language corpus resources.

• 2006

  Formation of the National Corpus of Polish consortium

  [1]National Corpus of Polishen.wikipedia.org

  Four previously separate corpus-building groups joined together: two institutes of the Polish Academy of Sciences, Polish Scientific Publishers PWN, and the University of Łódź’s computational and corpus linguistics department.

• Registration as a national research-and-development project

  [1]National Corpus of Polishen.wikipedia.org

  The initiative was registered as a research-and-development project of Poland’s Ministry of Science and Higher Education.

• Release of a balanced and annotated corpus resource

  [1]National Corpus of Polishen.wikipedia.org

  The project established a balanced subcorpus of about 300 million words and released a manually annotated corpus of about one million words under an open license.

• Online search access

  [2]National Corpus of Polish official websitenkjp.pl

  The corpus became accessible online through NKJP search services, including the Poliqarp interface.

A national scholarly reference resource for evidence-based research on Polish language use, combining broad textual coverage with searchable and annotated data.

• Linguistic corpus

  National Corpus of Polish

  The core NKJP resource is a large, multi-genre collection of Polish texts intended for empirical language research. It includes literary works, general newspapers, specialist publications, conversation transcripts, ephemeral material, and Internet texts. Its planned overall scale exceeded one billion words, with a balanced subcorpus of approximately 300 million words designed to offer more representative coverage. Researchers can use the data to examine meanings, grammatical functions, collocations, frequency, register, and change over time.

• Annotated language dataset

  Manually annotated corpus

  A manually annotated portion of approximately one million words was released under an open license. Annotation makes the material especially useful for detailed grammatical investigation and for computational applications such as tagging, parsing, and the development or evaluation of Polish-language processing systems.

• Corpus search engine

  Poliqarp

  Poliqarp provides query access to the corpus. Users can search for individual words, phrases, grammatical forms, and patterned sequences. Its functionality includes regular-expression-style searches and morphological expansion, allowing researchers to retrieve multiple forms of a lexical item or investigate structural combinations such as noun–adjective sequences.

• Corpus analysis platform

  PELCRA

  PELCRA combines data from three corpus collections associated with the participating institutions and provides an extensive environment for lexical and phraseological analysis. It supports flexible queries, morphological and spelling-related options, spreadsheet downloads, visual exploration of frequency or register patterns, and time-series analysis for words, phrases, and idioms.

1. 1National Corpus of Polishen.wikipedia.org
2. 2National Corpus of Polish official websitenkjp.pl