Dear Yan, Thank you for your email and for sharing your background with us. It's great to hear about your interest in the PersonalAIs project, and your experience in Music Emotion Recognition, NLP, and Full-Stack Development aligns perfectly with the goals of the project. The PersonalAIs project aims to develop an AI-powered agent for personalized music recommendations by leveraging natural language processing (NLP) to understand the user's emotional state and musical preferences. The system will generate and refine music playlists based on user input, utilizing the Spotify API for authentication and playlist management. It will also support real-time conversational modifications to playlists, allowing users to adjust mood, genre, and energy preferences during their interaction with the agent. Core technologies include: - NLP techniques for detecting user mood and preferences - Generative AI models (small LLMs) for dialogue generation - Integration with the Spotify API for user authentication and playlist management - A web-based chatbot UI for interacting with the agent - Real-time playlist modifications based on user feedback In terms of your inquiry about evaluation metrics, we are considering using both industry-standard recommendation evaluation metrics like NDCG (Normalized Discounted Cumulative Gain) and MRR (Mean Reciprocal Rank) and a more user-feedback-driven approach. A hybrid approach could help us balance objective performance metrics with subjective user satisfaction, and we would welcome your thoughts on how to best combine these methods. Looking forward to your continued interest and contributions to the project, and please feel free to share any further questions or ideas you may have. Best regards, Giannis Prokopiou & Thanos Aidinis Στις Δευ 17 Μαρ 2025 στις 6:41 π.μ., ο/η Yan Zhang <akira [ dot ] zhangy [ at ] gmail [ dot ] com> έγραψε: > Dear Giannis & Thanos, > > I hope you are having a good day. I am Yan Zhang, a master student in IT & > Cognition at Copenhagen University. I have been closely following the > discussions around the PersonalAIs project, and I am very interested in > contributing, and it aligns closely with my background. > > With my background in Music Emotion Recognition, NLP, and Full-Stack > Development, I have worked on several projects that align with the goals of > PersonalAIs: > > - Music Emotion Recognition: I conducted a comparative study on the > performance of unimodal (audio or lyrics) and multimodal (combined audio > and lyrics) models in Music Emotion Recognition (MER). > - NLP & Machine Learning: I have experience in text classification, > sentiment analysis, and author style recognition, using a machine > model with NLP techniques. > - Full-Stack Development: I've developed a responsive e-commerce > platform for yoga products, featuring real-time price calculations, user > authentication, and secure payment processing. Built RESTful APIs using > Express.js, designed the frontend with React.js, integrated Stripe API for > payments, and implemented MongoDB for data storage. > - Co-Pilot and Question Answering System – GraphRAG for Workflow > Understanding: I'm currently doing my thesis about GraphRAG and fine-tuning > SLM models, which could be useful in exploring how PersonalAIs can leverage > models for interactive recommendations. > > By integrating these experiences, I believe I can contribute effectively > to developing a personalized music recommendation system for the > PersonalAIs project as part of GSOC 2025. I am excited about this > opportunity and eager to explore how I can contribute. > > Additionally, I would like to ask about the evaluation criteria for the > project: > > - Does PersonalAIs have predefined evaluation standards? If not, do > you think it is necessary to establish them? > - Would you prefer to use industry-standard recommendation evaluation > metrics (e.g., NDCG, MRR), or would a more user-feedback-driven approach be > more suitable? > > I would greatly appreciate your insights on this. Looking forward to your > feedback! > > Best regards, > Yan Zhang > MSc IT & Cognition, University of Copenhagen > ---- > Λαμβάνετε αυτό το μήνυμα απο την λίστα: Λίστα αλληλογραφίας και συζητήσεων > που απευθύνεται σε φοιτητές developers \& mentors έργων του Google Summer > of Code - A discussion list for student developers and mentors of Google > Summer of Code projects., > https://lists.ellak.gr/gsoc-developers/listinfo.html > Μπορείτε να απεγγραφείτε από τη λίστα στέλνοντας κενό μήνυμα ηλ. > ταχυδρομείου στη διεύθυνση <gsoc-developers+unsubscribe [ at ] ellak [ dot ] gr>. >
---- Λαμβάνετε αυτό το μήνυμα απο την λίστα: Λίστα αλληλογραφίας και συζητήσεων που απευθύνεται σε φοιτητές developers \& mentors έργων του Google Summer of Code - A discussion list for student developers and mentors of Google Summer of Code projects., https://lists.ellak.gr/gsoc-developers/listinfo.html Μπορείτε να απεγγραφείτε από τη λίστα στέλνοντας κενό μήνυμα ηλ. ταχυδρομείου στη διεύθυνση <gsoc-developers+unsubscribe [ at ] ellak [ dot ] gr>.