CLARIN-PL
Not a member yet
504 research outputs found
Sort by
Unified Word Sense Disambiguation Inventory for Polish
No full textAn updated inventory for Polish Word Sense Disambiguation corpora. Includes an updated version of old sense-annotated treebank - Składnica (and, in the next version, an updated KPWr annotation), PolEval’s 2020 sense-annotated corpora named SH (Sherlock Holmes - The Adventure of the speckled band) and KPWr-100 (a new subset of 100 documents from KPWr), and Wikiselect which is a part of Wikipedia-based sense definitions from plWordNet 4.2. The sense annotations are compatible with plWordNet 4.2. The work is on progress and will be published officially soon
Cleaned Polish Oscar corpus (96M lines)
No full textCleaned Polish Oscar corpus (part: 96M lines, 3.49 GB). Data was prepared with a few cleaning heuristics:
- remove sentences shorter than
- remove non-polish sentences
- remove ungrammatical sentences
- perform sentence tokenization and save each sentence in a new line, after each document the new line was adde
Cleaned Polish Oscar corpus (128M above lines)
No full textCleaned Polish Oscar corpus (part: 128M above lines, 1.93 GB). Data was prepared with a few cleaning heuristics:
- remove sentences shorter than
- remove non-polish sentences
- remove ungrammatical sentences
- perform sentence tokenization and save each sentence in a new line, after each document the new line was adde
HerBERT Large Pre-trained on KGR10 Data
No full textHerBERT-large model fine-tuned on KGR10 data. The model was trained using DeepSpeed technology
MultiEmo: Multilingual, Multilevel, Multidomain Sentiment Analysis Corpus of Consumer Reviews
No full textMultiEmo, a new benchmark data set for the multilingual sentiment analysis task including 11 languages. The collection contains consumer reviews from four domains: medicine, hotels, products and university. The original reviews in Polish contained 8,216 documents consisting of 57,466 sentences. The reviews were manually annotated with sentiment at the level of the whole document and at the level of a sentence (3 annotators per element). We achieved a high Positive Specific Agreement value of 0.91 for texts and 0.88 for sentences. The collection was then translated automatically into English, Chinese, Italian, Japanese, Russian, German, Spanish, French, Dutch and Portuguese. MultiEmo is publicly available under a Creative Commons Attribution 4.0 International Licence.
More information: https://github.com/CLARIN-PL/multiemo
Citation:
@inproceedings{kocon2021multiemo,
title={Multiemo: Multilingual, multilevel, multidomain sentiment analysis corpus of consumer reviews},
author={Koco{\'n}, Jan and Mi{\l}kowski, Piotr and Kanclerz, Kamil},
booktitle={International Conference on Computational Science},
pages={297--312},
year={2021},
organization={Springer}
Mapping plWordNet 3.2 onto Linked Open Data - Manual Dataset
No full textThe mapping contains links to
- AGROVOC
- DDC
- DIGIZAURUS
- EUROVOC
- GEMET
- IATE
- ICD10
- KABA
- LCSH
- MESH
- RAMEAU
- STERNIK
- STW
- UDC
- UMLS
- WIKIPEDI
Big data language model stemmed in ARPA format
No full textBig data language model stemmed in ARPA format
Big data language model stemmed with BPE in ARPA format
No full textBig data language model stemmed with BPE in ARPA forma
Badanie infosfery Kronik Fallathanu - Inforex, materiały.
No full textKorpus, w którego skład wchodzą przykładowe teksty pozyskane ze strony Kroniki Fallathanu. Posłużyły one analizie najczęściej występujących anotacji, które pozwoliły na określenie struktury infosfery gry oraz skali dezinformacji
Big Data language model - STEMMED - RAW data
No full textBig data language model stemmed in RAW forma