The integration of Large Language Models (LLMs) into e-commerce platforms has significantly enhanced user experience through personalized recommendations and automated customer support. However, existing Retrieval-Augmented Generation (RAG) frameworks face challenges when applied to e-commerce product Question Answering (QA), such as handling extensive product catalogs, ensuring timely knowledge updates, and maintaining efficient retrieval performance. In this paper, we propose ItemRAG, a novel framework that combines RAG with item-based knowledge computing to address these challenges. ItemRAG decouples QA templates from specific products by leveraging a dynamic knowledge graph, enabling efficient updates and reducing the size of the knowledge base. The framework includes state analysis to capture user intent and context, grouped indexing for efficient retrieval, and knowledge computing to dynamically generate accurate answers. Experimental results demonstrate that decoupled-based ItemRAG significantly outperforms the Coupled-based RAG approaches (CoupledRAG) in retrieval accuracy and generation quality, achieving higher precision, recall, F1-score, and factual correctness. Our work highlights the efficacy of integrating the knowledge graph with RAG to enhance LLM-based e-commerce customer service systems.

Log parsing is indispensable for system maintenance, converting unstructured log data into structured formats (log templates) for further log compression and anomaly detection. The effectiveness of log parsing relies on the efficiency of two key processes: template extraction and log matching. Traditional methods, however, suffer from slow pairwise comparisons for template extraction and the tedious, non-scalable sequential approach for template matching. Our research has uncovered two opportunities for optimization based on two log locality characteristics: logs from the same template tend to cluster sequentially, and there is a limited variety of templates used within given timeframes. To exploit these opportunities, we propose the Multi-Logs Template Extraction (MLTE)-Cache framework. MLTE-Cache leverages the MLTE algorithm to enhance the efficiency of template extraction by grouping similar logs and processing them in batch mode. Furthermore, the framework utilizes a cache-assisted proximity matching algorithm to accelerate the log matching procedure. Through comprehensive experiments on open-source datasets, the MLTE-Cache framework has proven highly effective, maintaining a high level of accuracy while delivering a 37% improvement in efficiency.