PubAnnotation - a persistent and sharable corpus and annotation repository

摘要

There has been an active development of corpora and annotations in the BioNLP community. As those resources accumulate, a new issue arises about the reusability. As a solution to improve the reusability of corpora and annotations, we present PubAnnotation, a persistent and sharable repository, where various corpora and annotations can be stored together in a stable and comparable way. As a position paper, it explains the motivation and the core concepts of the repository and presents a prototype repository as a proof-of-concept.