SENTIMENT ANALYSIS AND DETERMINATION OF ASPECTS WITH RATINGS IN SOCIAL COMMENTS THROUGH TRAINED GENERATIVE MODELS

Rajabov , Jaloliddin, Matlatipov, Sanatbek

Techscience.uz - техника фанлари долзарб масалалри · 2025-yil

Annotatsiya

Typically, a review includes not only an overall rating but also ratings for several aspects and accompanying text. The rating is considered a numerical representation of the author's overall satisfaction. Although the number of reviews with aspect-specific ratings is increasing, there are still many reviews that only provide an overall rating. Extracting hidden aspect-related opinions from such reviews helps users quickly understand the gist without reading the entire text. This task mainly consists of two parts: identifying aspects and assigning ratings. Most existing studies cannot utilize the aspect ratings that have been increasingly available in recent years. In this article, we examine two artificial intelligence models that improve the efficiency of assigning aspect ratings to unseen reviews. Specifically, we look at sentiment words and aspect-level sentiment distributions that generate aspect ratings

Maqola ma’lumotlari
MualliflarRajabov , Jaloliddin, Matlatipov, Sanatbek
JurnalTechscience.uz - техника фанлари долзарб масалалри
Nashr sanasi2025-08-11
Jild3
Son5
Betlar41-50
TilO‘zbek
DOI10.47390/ts-v3i5y2025n7

Kalit so‘zlar

Uzbek language, sentiment analysis, text processing, aspect identification, O‘zbek tili, sentiment tahlil, matnlarni qayta ishlash, aspektlarni aniqlash

Ilmiy soha

Techscience.uz - техника фанлари долзарб масалалри jurnalidan boshqa maqolalar

Techscience.uz - техника фанлари долзарб масалалри — barcha maqolalar