Texts for text mining and Machine Learning (data)

A dataset that combines the title and abstracts from the lens dataset for text mining and ML.

data("texts")

Format

id

character

texts

character

Details

The titles and abstracts are divided into separate files and then joined. A unique id is constructed from the paperid and row number separated by "_". The result is a data frame compliant with the emerging TIF format favoured by quanteda, spacyr etc. paperids can be reconstructed for joins at a later stage using tidyr::separate(). For use with spacy in Python use e.g. jsonlite::stream_out(texts, file("flights.jsonl")).