چکیده مقاله
In recent years, Aspect based Sentiment Analysis ABSA has emerged as an essential yet challenging in the discernment of emotional content within textual data ABSA is a fine grained task and encompasses numerous applications across diverse domains, including social media platforms, product reviews, and cinematic critiques, thus establishing it as a compelling field of inquiry Numerous scholars are actively engaged in the development of increasingly sophisticated sentiment analysis frameworks for dealing with ABSA Presently, the cutting edge ABSA models implemented in the English language and other widely utilized spoken tongues predominantly rely on the fine tuning paradigm of generic pre trained language models LLMs ; however, this methodology has been overlooked in the context of the Persian language The application of pre trained LLMs to the Persian linguistic has been largely confined to coarse grained sentiment analysis or detecting sentiment of specified aspect rather than addressing the complete task of ABSA In this research paper, we explore the efficacy of DORNA Llama3 in facilitating ABSA and subsequently introduce an innovative approach for comprehensive end to end Persian ABSA The parameter efficient fine tuning has been utilized to maintain the integrity of the original model while concurrently mitigating computational expenses The proposed method has been evaluated on Pars ABSA dataset, demonstrating superior performance relative to the baseline approach
کلیدواژهها
نویسندگان
Zohre Karimi
Assistant Professor, School of Engineering, Damghan University, Damghan, Iran
Milad Foroughi
Computer Engineer, School of Engineering, Damghan University, Damghan, Iran
شیوه ارجاع
Karimi, Zohre and Foroughi, Milad,1403,Leveraging Dorna-Llama3 for Persian End-to-End Aspect-Based Sentiment Analysis,3nd International Conference and 8th National Conference on Computers, information technology and applications of artificial intelligence,Ahvaz
ارائهشده در
مجموعه مقالات سومین کنفرانس بین المللی و هشتمین کنفرانس ملی کامپیوتر، فناوری اطلاعات و کاربردهای هوش مصنوعی30 بهمن 1403 · اهواز