ARTIFICIAL INTELLIGENCE-BASED EMERGENCY MEDICAL CALL PROCESSING SYSTEM

Makharov, Kodirbek, Ismatillaev, Muzaffar

Techscience.uz - техника фанлари долзарб масалалри · 2026-yil

Annotatsiya

This paper presents a system for analyzing emergency medical calls using artificial intelligence technologies for dispatch services. The system is based on a Whisper-Small model fine-tuned on datasets FeruzaSpeech and Uzbek Speech Corpus with 121378 samples and hybrid NER algorithms. The study achieved a 14.7% WER in speech recognition, while medical symptoms, addresses and age indicators were extracted with 88% accuracy. A processing latency of 1.8 seconds enables real-time application. This solution serves to reduce dispatcher workload and enhance the speed of emergency medical response by automating data processing and analysis.

Maqola ma’lumotlari
MualliflarMakharov, Kodirbek, Ismatillaev, Muzaffar
JurnalTechscience.uz - техника фанлари долзарб масалалри
Nashr sanasi2026-03-25
Jild4
Son3
Betlar5-11
TilO‘zbek
DOI10.47390/ts-v4i3y2026n01

Kalit so‘zlar

Natural Language Processing, Speech-to-Text, Whisper Model, Named Entity Recognition, Data Structuring., tabiiy tilni qayta ishlash, nutqni matnga aylantirish, whisper modeli, nomlangan ma’lumotlarni aniqlash, ma’lumotlarni strukturalash.

Ilmiy soha

Techscience.uz - техника фанлари долзарб масалалри jurnalidan boshqa maqolalar

Techscience.uz - техника фанлари долзарб масалалри — barcha maqolalar