Foundations for an adaptive virtual agent: first steps towards developing an evolving, user-oriented, feedback-driven personal assistant capable of interacting with smart environments
Moretti, Marco (A.A. 2023/2024) Foundations for an adaptive virtual agent: first steps towards developing an evolving, user-oriented, feedback-driven personal assistant capable of interacting with smart environments. Tesi di Laurea in Machine learning, Luiss Guido Carli, relatore Giuseppe Francesco Italiano, pp. 74. [Master's Degree Thesis]
Full text for this thesis not available from the repository.
Abstract/Index
General artificial intelligence (AGI) vs. narrow AI. Assistants vs. agents. The goal of this research. Theoretical foundations of virtual agents. Definition and role of virtual agents in AI. Degrees of autonomy in agents. Role of feedback and adaptability. Interaction with the physical environment. Current approaches to building (generalist?) Virtual agents. A survey of solutions. The role of large language models (LLMs). Retrieval-augmented generation (RAG) systems. Agent collaboration and interaction. Privacy vs. cost efficiency in LLM backends. Building virtual agents driven by user feedback. User and community feedback loops. Interactive learning. Memory and knowledge management for agents. Short-term vs. long-term memory in LLMs. Neo4j and graph databases for agent memory. Vector embeddings for long-term memory. Architectural design considerations. Speech-to-text (STT) backend options. Text-to-speech (TTS) backend options. Large language models (LLMs) backend. Memory management systems. Reverse proxy and API management. Frontend and user interface (dashboard). Smart home and environment control system. Interaction flow.
References
Bibliografia: pp. 71-73.
| 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: | Machine learning |
| Thesis Supervisor: | Italiano, Giuseppe Francesco |
| Thesis Co-Supervisor: | Coppa, Emilio |
| Academic Year: | 2023/2024 |
| Session: | Autumn |
| Deposited by: | Alessandro Perfetti |
| Date Deposited: | 09 Jun 2025 10:18 |
| Last Modified: | 09 Jun 2025 10:18 |
| URI: | https://tesi.luiss.it/id/eprint/42345 |
Downloads
Downloads per month over past year
Repository Staff Only
 |
View Item |
