Σχεδιασμός και υλοποίηση ευφυούς συστήματος επαγγελματικού προσανατολισμού βασισμένου σε RAG και open-source LLMs
Design and implementation of an intelligent career guidance system based on RAG and open-source LLMs
View/
Keywords
Retrieval-Augmented Generation (RAG) ; Large Language Models (LLMs) ; Career Guidance ; RIASEC Theory ; LangChain ; ChromaDB ; AI in Education ; EducationAbstract
In this thesis, we will design, develop, and evaluate an intelligent conversational agent to provide career guidance through using Retrieval-Augmented Generation (RAG) and large-scale Open-Source Large Language Models (LLMs). Developed using Python for the Backend and HTML/CSS/JavaScript for the Frontend, our primary purpose is to assist people in providing evidence-based personalized career advice in Greek that reduces LLM hallucinations by integrating a separate knowledge base.
The system we created uses a hybridised data architecture that integrates a MySQL relational database for maintaining structured user profiles and psychometric data based on John Holland's RIASEC model with a ChromaDB vector database to support unstructured information retrieval. The implementation of the system is based on the LangChain framework and Groq Cloud's infrastructure using state-of-the-art models such as GPT-OSS 120B and Llama 3.3 70B. We emphasised the use of the Recursive Character Chunking strategy and applying multilingual embeddings to enhance the efficacy of semantic search.
Postgraduate Studies Programme
Πληροφοριακά Συστήματα και ΥπηρεσίεςDepartment
Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών ΣυστημάτωνNumber of pages
105Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα