Lorenzo Di Rosa
An LLM-based Multi-Agent Pipeline for Academic Literature Review and Summarization.
Rel. Alessio Sacco, Guido Marchetto. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Informatica (Computer Engineering), 2026
|
Preview |
PDF (Tesi_di_laurea)
- Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (3MB) | Preview |
Abstract
In the modern academic landscape, the sheer volume of available data has transformed the nature of research from a quest for information into a battle against information overload. Compared to the past, where research was limited by the accessibility of physical archives, today’s scholars must navigate an immense, decentralized "sea of information." This shift has made the ability to synthesize and verify data more critical than ever before. While Large Language Models (LLMs) have emerged as potential solutions for document summarization, they remain inherently unreliable. Current mainstream tools, such as ChatGPT Scholar, often suffer from hallucinations, opaque reasoning, and a lack of verifiable accuracy.
To address these limitations, this thesis proposes a specialized Multi-Agent System (MAS) pipeline
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Corso di laurea
Classe di laurea
Aziende collaboratrici
URI
![]() |
Modifica (riservato agli operatori) |
