In legal informatics research, decision support systems can be a valuable tool for practitioners facing a growing volume of data. An expert system based on information retrieval and a recommender system can benefit from the application of Large Language Models to improve the quality of results. This paper proposes a general framework based on Retrieval-Augmented Generation for addressing integrated recommenda- tion systems with generative models in public procurement. Moreover, we addressed a practical application by adopting real datasets in the legal domain. To illustrate the feasibility of the approach, a proof-of-concept has been presented in the context of public procurement management within an Italian case study. The study and evaluation phases have been supervised by domain experts in the legal field to ensure robust analysis and relevance.
Large Language Models and Recommendation Systems: A Proof-of-Concept Study on Public Procurements
Nai, Roberto;Sulis, Emilio;Fatima, Ishrat;Meo, Rosa
2024-01-01
Abstract
In legal informatics research, decision support systems can be a valuable tool for practitioners facing a growing volume of data. An expert system based on information retrieval and a recommender system can benefit from the application of Large Language Models to improve the quality of results. This paper proposes a general framework based on Retrieval-Augmented Generation for addressing integrated recommenda- tion systems with generative models in public procurement. Moreover, we addressed a practical application by adopting real datasets in the legal domain. To illustrate the feasibility of the approach, a proof-of-concept has been presented in the context of public procurement management within an Italian case study. The study and evaluation phases have been supervised by domain experts in the legal field to ensure robust analysis and relevance.File | Dimensione | Formato | |
---|---|---|---|
nldb_2024_Large Language Models and Recommendation Systems.pdf
Accesso aperto
Dimensione
26.16 MB
Formato
Adobe PDF
|
26.16 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.