LlamaRec-LKG-RAG: A Single-Pass, Learnable Knowledge Graph-RAG Framework for LLM-Based Ranking
Vahid Azizi, Fatemeh Koochaki · 2025
Read LlamaRec-LKG-RAG: A Single-Pass, Learnable Knowledge Graph-RAG Framework for LLM-Based Ranking by Vahid Azizi, Fatemeh Koochaki (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-07449.txt.
Description
Vahid Azizi and Fatemeh Koochaki enrich language-model recommendations with personal paths through a knowledge graph. Standard retrieval for recommender prompts leans on flat similarity and misses the web of relations among users, items, and traits. Their framework extends the LlamaRec ranker with a learnable graph layer.
In their 2025 paper, a light user preference module picks the relation paths that matter for each person inside a graph built from behavior and item metadata. Those personal subgraphs fold into the prompt of a fine-tuned Llama-2 model, so ranking happens in a single pass. On movie and beauty product data sets, the graph-fed ranker beats its base on standard ranking scores, with clearer reasons to show.
Catalog record
| CREATOR | Vahid Azizi, Fatemeh Koochaki |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-07449