MLRS Corpora

Corpora are available from our corpus server and require a username and password, which can be created online. You can also download the corpora from here.

Korpus Malti: General Corpora of Maltese
New!
Korpus Malti v4.2 (2024)

Corpus of ca. 470 million tokens, in various genres.
Annotated with part of speech tags, lemmas and morphological roots.
Korpus Malti v3.0 (2016)

Corpus of ca. 250 million tokens, in various genres.
Annotated with part of speech tags, lemmas and morphological roots.
Korpus Malti v2.0 (Beta)

Corpus of ca. 130 million tokens, in various genres. Annotated with part of speech tags.
Further info

The Korpus Malti v3.0 (2016) is tagged with the Maltese Tagset v3.0, developed by Slavomir Céplö and Albert Gatt. Tagging was carried out using the SVMTool, trained on a manually tagged corpus of ca. 28,000 tokens, with an accuracy of ca. 97%.

Learner Corpora
Corpus of Learner English in Malta (CLEM)

Corpus of ca. 1 million tokens, consisting of English essays by students. These essays were sampled from examination scripts by Maltese students written during the 2011-2013 examinations sessions of the Matriculation and Secondary Education Certificate (MATSEC) Examinations Board.
The corpus is stratified by gender, school type, candidate's region of residence, date of birth and mark/grade. Tokens are annotated with part of speech, lemma and orthographic errors.
Further info

CLEM Corpora are an outcome of the CLEM Project, a collaboration between the Institute of Linguistics and Language Technology and the Centre for English Language Proficiency.

The POS tagging for these corpora was carried out using the Penn Treebank Tagset for English and the POS Tagger distributed as part of the Python NLTK library.