license: openrail++
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: 0-No
dtype: float64
- name: 6-Title
dtype: large_string
- name: 9-Full_text
dtype: large_string
- name: 10-Topic-general
dtype: large_string
- name: 16-Label
dtype: large_string
splits:
- name: train
num_bytes: 57329380
num_examples: 1048207
download_size: 8305806
dataset_size: 57329380
Dataset description
The corpus for the automatic identification of false information
The corpus for the automatic identification of false information comprises linguistic resources required for developing and advancing AI-based solutions capable of automatically detecting misleading information in online media texts. The total size of the corpus is 5,162 annotated entries. The level of misleading information was determined by consensus between two professional annotators. Metadata categories include: article title, article text, domain label, and a label assessing the level of misleading information.
MELAGINGOS INFORMACIJOS AUTOMATINIO IDENTIFIKAVIMO TEKSTYNAS
Melagingos informacijos automatinio identifikavimo tekstynas apima lingvistinius resursus, reikalingus kuriant ir vystant dirbtinio intelekto technologijomis grįstus sprendimus, kurie gebėtų automatiškai atpažinti klaidinančią informaciją interneto žiniasklaidos medijų tekstuose. Bendras Tekstyno dydis yra 5 162 anotuoti įrašai. Klaidinančios informacijos lygmens įvertį bendru sutarimu nustatė 2 (du) profesionalūs vertintojai. Metaduomenų kategorijos: straipsnio pavadinimas, straipsnio tekstas, srities žyma, klaidinančios informacijos lygį vertinanti žyma.
Acknowledgement
NextGenerationEU/Naujos kartos Lietuva Project code:02-104-P-0001 Project name:Development of Corpus for Automatic Identification of False Information