Adaptive reasoning conversation recommendation system using graph (ARCR) | |
| Author | Todsavad Tangtortan |
| Call Number | AIT Thesis no.DSAI-24-07 |
| Subject(s) | Recommender systems (Information filtering) Information visualization Neural networks (Computer science) Graph theory--Data processing |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Data Science and Artificial Intelligence |
| Publisher | Asian Institute of Technology |
| Abstract | Typically, one of current recommendation system’s methods rely on generating a paths from users to items to recommend using past user behavior which train entity embeddings or Graph Neural Nerwork (GNNs). This may significant challenge for new users which have lack of interaction history. Multi-round Conversational Recommendations address the cold-start by asking users questions to gather more in- formation about their preferences through conversation. Through questioning, it deduces the attributes users prefer or do not prefer. However, these techniques only consider preferred attributes in dependently and ignore the explicit reasoning behind why these items are recommended. We propose a framework that first identifies a user’s preferences through conversations, distinguishing positive and negative preferences to build apersonalizedsubgraphprofile. Next, we align this profile with the top-k historical preferences of other users to identify the most likely preferences that correspond with this profile, constructing a reasoned graph for recommending items. Finally, we refine this graph by trimming paths associated with negative preferences or items. This method not only accommodates users with limited initial interaction data but also mitigates cold-start even containing only negative preferences, thereby offering the accuracy and relevance of recommendations for new users. |
| Year | 2024 |
| Type | Thesis |
| School | School of Engineering and Technology |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Data Science and Artificial Intelligence (DSAI) |
| Chairperson(s) | Chaklam Silpasuwanchai; |
| Examination Committee(s) | Chantri Polprasert;Attaphongse Taparugssanagorn; |
| Scholarship Donor(s) | Royal Thai Government; |
| Degree | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Data Science and Artificial Intelligence |