Standard
Entity-Aware Sentiment Analysis for Uzbek Social Media Using XLM-RoBERTa NER and LSTM. / Saidov, Bobur R.; Fayzullaeva, Zarnigor; Tursunov, Ulugbek A. et al.
Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026. Institute of Electrical and Electronics Engineers Inc., 2026. (Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026).
Research output: Chapter in Book/Report/Conference proceeding › Chapter › Research › peer-review
Harvard
Saidov, BR, Fayzullaeva, Z, Tursunov, UA, Narmuratov, Z, Egamova, SD & Gulbakhar, BA 2026,
Entity-Aware Sentiment Analysis for Uzbek Social Media Using XLM-RoBERTa NER and LSTM. in
Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026. Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026, Institute of Electrical and Electronics Engineers Inc.
https://doi.org/10.1109/USBEREIT70063.2026.11580443APA
Saidov, B. R., Fayzullaeva, Z., Tursunov, U. A., Narmuratov, Z., Egamova, S. D., & Gulbakhar, B. A. (2026).
Entity-Aware Sentiment Analysis for Uzbek Social Media Using XLM-RoBERTa NER and LSTM. In
Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026 (Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026). Institute of Electrical and Electronics Engineers Inc..
https://doi.org/10.1109/USBEREIT70063.2026.11580443Vancouver
Saidov BR, Fayzullaeva Z, Tursunov UA, Narmuratov Z, Egamova SD, Gulbakhar BA.
Entity-Aware Sentiment Analysis for Uzbek Social Media Using XLM-RoBERTa NER and LSTM. In Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026. Institute of Electrical and Electronics Engineers Inc. 2026. (Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026). doi: 10.1109/USBEREIT70063.2026.11580443
Author
Saidov, Bobur R. ; Fayzullaeva, Zarnigor ; Tursunov, Ulugbek A. et al. /
Entity-Aware Sentiment Analysis for Uzbek Social Media Using XLM-RoBERTa NER and LSTM. Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026. Institute of Electrical and Electronics Engineers Inc., 2026. (Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026).
BibTeX
@inbook{c6fb4a26dd9e4609a35e2d8805c2fa14,
title = "Entity-Aware Sentiment Analysis for Uzbek Social Media Using XLM-RoBERTa NER and LSTM",
abstract = "This paper studies how named entities influence sentiment polarity in Uzbek social media and proposes an entity-aware sentiment analysis pipeline. We collected 10,000 Uzbek posts from Telegram, Instagram, and Facebook (January-June 2025) and annotated three entity types (Person, Brand, Geolocation) in a subset of 4,600 posts. Named entities are extracted with an adapted XLM-RoBERTa NER model, while sentiment is predicted by an LSTM classifier with Uzbek Word2Vec embeddings. To account for the different impact of entity types, we introduce an entity-weighted adjustment that calibrates sentiment scores using empirically estimated coefficients. On the held-out test split, the proposed approach achieves strong sentiment performance, reaching Test Accuracy=0.84 and Macro-F1=0.83 (Macro Precision=0.83, Macro Recall = 0.82). Statistical validation confirms significant differences across entity categories (ANOVA) and a positive association between entity mentions and model confidence (Pearson r=0.82). The approach supports practical analytics scenarios such as brand reputation monitoring, public communication tracking, and tourism feedback analysis for Uzbek-language data.",
keywords = "NLP, Uzbek language, named objects, sentiment analysis, social networks, Моделирование, Управление брендом, Печать, Социальные сети (онлайн), Тестирование, Анализ настроений, Дисперсионный анализ, Разметка, Многоязычность, Долговременная кратковременная память, именованные объекты, анализ настроений, узбекский язык, НЛП, социальные сети",
author = "Saidov, {Bobur R.} and Zarnigor Fayzullaeva and Tursunov, {Ulugbek A.} and Zayniddin Narmuratov and Egamova, {Shokhida D.} and Gulbakhar, {Begmuratova A.}",
note = "B. R. Saidov, Z. Fayzullaeva, U. A. Tursunov, Z. Narmuratov, S. D. Egamova and B. A. Gulbakhar, {"}Entity-Aware Sentiment Analysis for Uzbek Social Media Using XLM-RoBERTa NER and LSTM,{"} 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology (USBEREIT), Yekaterinburg, Russian Federation, 2026, pp. 1-4, doi: 10.1109/USBEREIT70063.2026.11580443. keywords: {Modeling;Brand management;Printing;Social networking (online);Testing;Sentiment analysis;Analysis of variance;Labeling;Multilingual;Long short term memory;named objects;sentiment analysis;Uzbek language;NLP;social networks}, ",
year = "2026",
doi = "10.1109/USBEREIT70063.2026.11580443",
language = "English",
isbn = "9798319541772",
series = "Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
booktitle = "Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026",
address = "United States",
}
RIS
TY - CHAP
T1 - Entity-Aware Sentiment Analysis for Uzbek Social Media Using XLM-RoBERTa NER and LSTM
AU - Saidov, Bobur R.
AU - Fayzullaeva, Zarnigor
AU - Tursunov, Ulugbek A.
AU - Narmuratov, Zayniddin
AU - Egamova, Shokhida D.
AU - Gulbakhar, Begmuratova A.
N1 - B. R. Saidov, Z. Fayzullaeva, U. A. Tursunov, Z. Narmuratov, S. D. Egamova and B. A. Gulbakhar, "Entity-Aware Sentiment Analysis for Uzbek Social Media Using XLM-RoBERTa NER and LSTM," 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology (USBEREIT), Yekaterinburg, Russian Federation, 2026, pp. 1-4, doi: 10.1109/USBEREIT70063.2026.11580443. keywords: {Modeling;Brand management;Printing;Social networking (online);Testing;Sentiment analysis;Analysis of variance;Labeling;Multilingual;Long short term memory;named objects;sentiment analysis;Uzbek language;NLP;social networks},
PY - 2026
Y1 - 2026
N2 - This paper studies how named entities influence sentiment polarity in Uzbek social media and proposes an entity-aware sentiment analysis pipeline. We collected 10,000 Uzbek posts from Telegram, Instagram, and Facebook (January-June 2025) and annotated three entity types (Person, Brand, Geolocation) in a subset of 4,600 posts. Named entities are extracted with an adapted XLM-RoBERTa NER model, while sentiment is predicted by an LSTM classifier with Uzbek Word2Vec embeddings. To account for the different impact of entity types, we introduce an entity-weighted adjustment that calibrates sentiment scores using empirically estimated coefficients. On the held-out test split, the proposed approach achieves strong sentiment performance, reaching Test Accuracy=0.84 and Macro-F1=0.83 (Macro Precision=0.83, Macro Recall = 0.82). Statistical validation confirms significant differences across entity categories (ANOVA) and a positive association between entity mentions and model confidence (Pearson r=0.82). The approach supports practical analytics scenarios such as brand reputation monitoring, public communication tracking, and tourism feedback analysis for Uzbek-language data.
AB - This paper studies how named entities influence sentiment polarity in Uzbek social media and proposes an entity-aware sentiment analysis pipeline. We collected 10,000 Uzbek posts from Telegram, Instagram, and Facebook (January-June 2025) and annotated three entity types (Person, Brand, Geolocation) in a subset of 4,600 posts. Named entities are extracted with an adapted XLM-RoBERTa NER model, while sentiment is predicted by an LSTM classifier with Uzbek Word2Vec embeddings. To account for the different impact of entity types, we introduce an entity-weighted adjustment that calibrates sentiment scores using empirically estimated coefficients. On the held-out test split, the proposed approach achieves strong sentiment performance, reaching Test Accuracy=0.84 and Macro-F1=0.83 (Macro Precision=0.83, Macro Recall = 0.82). Statistical validation confirms significant differences across entity categories (ANOVA) and a positive association between entity mentions and model confidence (Pearson r=0.82). The approach supports practical analytics scenarios such as brand reputation monitoring, public communication tracking, and tourism feedback analysis for Uzbek-language data.
KW - NLP
KW - Uzbek language
KW - named objects
KW - sentiment analysis
KW - social networks
KW - Моделирование
KW - Управление брендом
KW - Печать
KW - Социальные сети (онлайн)
KW - Тестирование
KW - Анализ настроений
KW - Дисперсионный анализ
KW - Разметка
KW - Многоязычность
KW - Долговременная кратковременная память
KW - именованные объекты
KW - анализ настроений
KW - узбекский язык
KW - НЛП
KW - социальные сети
UR - https://www.mendeley.com/catalogue/71f507e2-0446-3c5e-833d-945c63a98699/
UR - https://www.scopus.com/pages/publications/105044705736
U2 - 10.1109/USBEREIT70063.2026.11580443
DO - 10.1109/USBEREIT70063.2026.11580443
M3 - Chapter
SN - 9798319541772
T3 - Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026
BT - Proceedings - 2026 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2026
PB - Institute of Electrical and Electronics Engineers Inc.
ER -
