Persian Text Summarization via Fine Tuning mT۵Transformer
محل انتشار: اولین کنفرانس بین المللی و ششمین کنفرانس ملی کامپیوتر، فناوری اطلاعات و کاربردهای هوش مصنوعی
سال انتشار: 1401
نوع سند: مقاله کنفرانسی
زبان: انگلیسی
مشاهده: 124
فایل این مقاله در 6 صفحه با فرمت PDF قابل دریافت می باشد
- صدور گواهی نمایه سازی
- من نویسنده این مقاله هستم
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
CEITCONF06_058
تاریخ نمایه سازی: 26 خرداد 1402
چکیده مقاله:
Nowadays, one of the main challenges in the world ofbig data is finding important information from a large text. Atany moment, large volumes of news from different newsagencies are being broadcasted, so reading all this news willrequire spending a lot of time. One of the ways to deal with thisproblem is to summarize the text automatically. Textsummarization is one of the most important and challengingtasks in the field of natural language processing. Summarizingtext means converting the relevant text into a concise form sothat its size is reduced in terms of the number of words,without changing the overall meaning of the text. In recentyears, transformers have been very successful in naturallanguage processing tasks, especially text summarization. Inthis research, we have trained two pre-trained transformers onPersian news texts, mT۵-base and mT۵-small, and evaluatedtheir performance in summarizing Persian texts
کلیدواژه ها:
نویسندگان
Vahid Nejad Mahmoodabadi
Computer Engineering Department Faculty of EngineeringShahid Bahonar University of KermanKerman, Iran
Fahimeh Ghasemian
Computer Engineering Department Faculty of EngineeringShahid Bahonar University of KermanKerman, Iran