ivanhe/opencap
The NYU Open CAP dataset consists of 50 patents related to the technology field of speech recognition, in which citations to scientific articles are annotated with CoNLL style B/I/O tags.
Shift format ECMAScript 5 AST tooling
Public repository record indexed from GitHub. Explore verified star velocity metrics, source code repositories, and curated developer tool directories across the GitHubRepo ecosystem.
The NYU Open CAP dataset consists of 50 patents related to the technology field of speech recognition, in which citations to scientific articles are annotated with CoNLL style B/I/O tags.
中文自然语言处理工具包 Toolkit for Chinese natural language processing (formerly FudanNLP)
AIDA Named Entity Disambiguation by the Databases and Information Systems Group at the Max Planck Institute for Informatics.
Public repository.