Semantic XPath: Structured Agentic Memory Access for Conversational AI

2026-03-01

AI Summary

Current dialogue AI systems are constrained in long-term tasks by limited context windows or the absence of structured memory mechanisms. This work proposes Semantic XPath, a tree-structured memory module that integrates semantic path-based querying with retrieval-augmented generation (RAG) to enable efficient and precise memory access and updates. By uniquely combining structured memory organization with a semantic XPath query mechanism, the approach substantially overcomes the limitations of conventional flat memory models. Experimental results demonstrate a 176.7% performance improvement over flat RAG baselines while consuming only 9.1% of the token budget, thereby validating the effectiveness and superiority of structured memory for long-term dialogue scenarios.

Technology Category

• Natural Language Processing: Conversational AI/Dialog Systems
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Conversational AI (ConvAI) agents increasingly maintain structured memory to support long-term, task-oriented interactions. In-context memory approaches append the growing history to the model input, which scales poorly under context-window limits. RAG-based methods retrieve request-relevant information, but most assume flat memory collections and ignore structure. We propose Semantic XPath, a tree-structured memory module to access and update structured conversational memory. Semantic XPath improves performance over flat-RAG baselines by 176.7% while using only 9.1% of the tokens required by in-context memory. We also introduce SemanticXPath Chat, an end-to-end ConvAI demo system that visualizes the structured memory and query execution details. Overall, this paper demonstrates a candidate for the next generation of long-term, task-oriented ConvAI systems built on structured memory.

Problem

• Conversational AI
• structured memory
• memory access
• RAG
• context window

Innovation

• Semantic XPath
• structured memory
• conversational AI
• tree-structured retrieval
• memory efficiency