Semantic XPath: Structured Agentic Memory Access for Conversational AI

TL;DR

Semantic XPath introduces a tree-structured memory module for conversational AI, significantly enhancing long-term memory access efficiency and performance over flat memory approaches, and is demonstrated through an interactive demo system.

Contribution

The paper presents Semantic XPath, a novel tree-structured memory access method for ConvAI, improving memory efficiency and performance over existing flat-memory retrieval methods.

Findings

Improves performance over flat-RAG by 176.7%.

Uses only 9.1% of tokens compared to in-context memory.

Demonstrates a visualized ConvAI system with structured memory.

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.