A Multi-Source Synthetic Dataset for Uzbek Sentiment Analysis, Named Entity Recognition, and Normalization

Published: 2 January 2026| Version 5 | DOI: 10.17632/khnmnp4t7v.5
Contributor:
Bobur Saidov

Description

This repository provides a multi-source synthetic Uzbek dataset for (i) sentiment classification (Positive/Neutral/Negative) and (ii) named entity recognition with PER/LOC/ORG/DATE labels, plus auxiliary resources for emoji-aware modeling and text normalization. The main file contains 10,000 unique sentences with aligned entity spans (surface forms + types) and an emoji-aware score in [-1,1]. Emoji usage is source-dependent (news ~15%, social ~75%, dialog ~55%) to better reflect real communication styles. All data were generated programmatically from rule-based templates and lexicons; no copyrighted or real user content was used. Primary formats are CSV and JSONL (XLSX provided only for convenience).

Files

Steps to reproduce

Steps to reproduce (v6, short): Load lexical resources: Uzbek personal names (12,000), geographic names (122), informal→normalized pairs (1,292), emoji polarity lexicon (2,000), emoji combinations (16,000), ASCII emoji variants (104), plus ORG/DATE patterns used in templates. Generate sentences: choose a template (Positive/Neutral/Negative) and a source style (synthetic_news, synthetic_social, synthetic_dialog); sample fillers; insert entities (PER/LOC/ORG/DATE) and apply simple Uzbek morphological rules (case endings, basic agreement). Emoji cues (optional): sample emoji/combos by polarity rules and insert with source-dependent frequency (lower in news, higher in social/dialog). Normalize & QC: apply informal→normalized replacements when needed; enforce UTF-8; remove duplicates; validate JSON arrays in CSV (entities, entity_type) have equal length and match the text. Export: primary formats are CSV and JSONL (XLSX convenience). Final schema: id, text, polarity, entities, entity_type, emoji_score, token_count, source. Deterministic rules + fixed random seeds allow consistent regeneration across machines.

Institutions

  • Al Xorazmiy nomidagi Urganch davlat universiteti
  • Novosibirskij gosudarstvennyj universitet

Categories

Computer Science, Artificial Intelligence, Computational Linguistics, Data Science, Natural Language Processing, Machine Learning, Human Language Resources, Text Mining, Sentiment Analysis, Cross-Lingual Learning

Licence