Developing of Mini-Scientific Chatbot based on Transformers and LLMs
| dc.contributor.advisor | Said, Gadri | |
| dc.contributor.author | Roqiya, Allal | |
| dc.contributor.author | Faten Aichaoui | |
| dc.date.accessioned | 2026-07-07T12:59:38Z | |
| dc.date.issued | 2026-06-10 | |
| dc.description.abstract | This thesis presents MedSci-GPT, a lightweight multilingual scientific chatbot using a hybrid architecture: GPT-4o API (primary) and fine-tuned Phi-2 (fallback). LoRA (rank 8, alpha 16) was applied on scientific QA datasets. A web app was built with PHP, MySQL, HTML/CSS/JS supporting authentication, conversation management, and Arabic/English/French. Evaluation shows Phi-2 achieves 76% accuracy (0.8s) and GPT-4o 82% (1.8s), proving lightweight scientific assistants run on standard hardware. | |
| dc.identifier.uri | https://depot.univ-msila.dz/handle/123456789/48862 | |
| dc.language.iso | en | |
| dc.publisher | University of M'sila | |
| dc.subject | Large Language Models | |
| dc.subject | Scientific Chatbot | |
| dc.subject | Fine-tuning | |
| dc.subject | LoRA | |
| dc.subject | GPT-4o | |
| dc.subject | Phi- 2 | |
| dc.subject | Multilingual | |
| dc.subject | Arabic NLP | |
| dc.subject | Question Answering | |
| dc.subject | Transformer | |
| dc.title | Developing of Mini-Scientific Chatbot based on Transformers and LLMs | |
| dc.type | Thesis |