UzLex Dataset: Core Uzbek Lemmas, Sentiment Lexicon, and PER/ORG/LOC/ABBR Gazetteers with Quality Control

UzLex is a curated multi-table Uzbek lexical dataset comprising (i) a core lemma lexicon with POS labels and quality-control flags, (ii) a sentiment lexicon with polarity labels and uncertainty flags, and (iii) gazetteers for NER (PER/ORG/LOC/ABBR). The release includes audit artifacts, quality summaries, and high-precision subsets for strict downstream use