Trendminer Hungarian Processing Pipeline (trendminer-hunlp): a suite of scripts that perform Hungarian NLP processing steps (tokenization, pos-tagging, morphological analysis, lemmatization) by extending existing tools (huntoken, hunmorph, hunpos) to be able to deal with some of the challenges presented by the special language of social media messages, which differs from the domain of standard language (generally newswire) texts that were used to develop and train the exising tools.
Trendminer Project:
Author: Márton Miháltz [email protected]