SPIRIT: Single pipeline & intelligent reporting for integrated tracking: an end-to-end framework for brand monitoring integrating semantic filtering and GPT-driven reporting

D'Onghia, Edoardo (A.A. 2023/2024) SPIRIT: Single pipeline & intelligent reporting for integrated tracking: an end-to-end framework for brand monitoring integrating semantic filtering and GPT-driven reporting. Tesi di Laurea in Data science in action, Luiss Guido Carli, relatore Alessio Martino, pp. 59. [Master's Degree Thesis]

[img]
Preview
PDF (Full text)
Download (701kB) | Preview

Abstract/Index

Problem statement. Objectives of SPIRIT. Background and related works. Brand monitoring. Generative AI. Sentiment analysis. Topic modeling. Relevance filtering via embeddings. Integration for modern brand monitoring. Literature overview. Positioning of SPIRIT. System architecture. High-level workflow. Data flow. Caching strategy. Implementation details. Technologies and libraries. Scrapers. Relevance filtering with embeddings. NLP modules. GPT-driven reporting. PDF generation. Evaluation and experimental results. Use case and interface. Scraping validation. Sentiment analysis per source. Il Giornale word frequencies (negative). Topic modeling and named entity recognition. GPT commentary and report generation. Qualitative validation against real events. Performance and overhead. The growing importance of brand monitoring. Towards an end-to-end pipeline. Data, AI and the role of human oversight. Modular, extensible architecture. Practical takeaways and broader implications.

References

Bibliografia: p. 59.

Thesis Type: Master's Degree Thesis
Institution: Luiss Guido Carli
Degree Program: Master's Degree Programs > Master's Degree Program in Data Science e Management (LM-91)
Chair: Data science in action
Thesis Supervisor: Martino, Alessio
Thesis Co-Supervisor: Italiano, Giuseppe Francesco
Academic Year: 2023/2024
Session: Extraordinary
Deposited by: Alessandro Perfetti
Date Deposited: 10 Jul 2025 12:14
Last Modified: 10 Jul 2025 12:14
URI: https://tesi.luiss.it/id/eprint/42875

Downloads

Downloads per month over past year

Repository Staff Only

View Item View Item