Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RUVA: Personalized Transparent On-Device Graph Reasoning

Gabriele Conte, Alessio Mattiace, Gianni Carmosino, Potito Aghilar, Giovanni Servedio, Francesco Musicco, Vito Walter Anelli, Tommaso Di Noia, Francesco Maria Donini

TL;DR

Ruva tackles privacy and explainability gaps in retrieval-based personalization by replacing opaque vector search with a Glass Box on-device Personal Knowledge Graph. It introduces EpisTwin, a Type 3 Neuro-Symbolic framework where schema-aware LLM-driven construction populates a PKG and a GraphRAG runtime performs $N$-hop topological reasoning grounded in explicit graph nodes. It enables deterministic deletion of memories via SQL cascade, supporting the Right to be Forgotten, and supports multimodal data via Chain-of-Experts ingestion. On-device evaluation on mobile shows feasible latency and strong cross-model agreement, suggesting practical viability for private, auditable personal AI assistants.

Abstract

The Personal AI landscape is currently dominated by "Black Box" Retrieval-Augmented Generation. While standard vector databases offer statistical matching, they suffer from a fundamental lack of accountability: when an AI hallucinates or retrieves sensitive data, the user cannot inspect the cause nor correct the error. Worse, "deleting" a concept from a vector space is mathematically imprecise, leaving behind probabilistic "ghosts" that violate true privacy. We propose Ruva, the first "Glass Box" architecture designed for Human-in-the-Loop Memory Curation. Ruva grounds Personal AI in a Personal Knowledge Graph, enabling users to inspect what the AI knows and to perform precise redaction of specific facts. By shifting the paradigm from Vector Matching to Graph Reasoning, Ruva ensures the "Right to be Forgotten." Users are the editors of their own lives; Ruva hands them the pen. The project and the demo video are available at http://sisinf00.poliba.it/ruva/.

RUVA: Personalized Transparent On-Device Graph Reasoning

TL;DR

Ruva tackles privacy and explainability gaps in retrieval-based personalization by replacing opaque vector search with a Glass Box on-device Personal Knowledge Graph. It introduces EpisTwin, a Type 3 Neuro-Symbolic framework where schema-aware LLM-driven construction populates a PKG and a GraphRAG runtime performs -hop topological reasoning grounded in explicit graph nodes. It enables deterministic deletion of memories via SQL cascade, supporting the Right to be Forgotten, and supports multimodal data via Chain-of-Experts ingestion. On-device evaluation on mobile shows feasible latency and strong cross-model agreement, suggesting practical viability for private, auditable personal AI assistants.

Abstract

The Personal AI landscape is currently dominated by "Black Box" Retrieval-Augmented Generation. While standard vector databases offer statistical matching, they suffer from a fundamental lack of accountability: when an AI hallucinates or retrieves sensitive data, the user cannot inspect the cause nor correct the error. Worse, "deleting" a concept from a vector space is mathematically imprecise, leaving behind probabilistic "ghosts" that violate true privacy. We propose Ruva, the first "Glass Box" architecture designed for Human-in-the-Loop Memory Curation. Ruva grounds Personal AI in a Personal Knowledge Graph, enabling users to inspect what the AI knows and to perform precise redaction of specific facts. By shifting the paradigm from Vector Matching to Graph Reasoning, Ruva ensures the "Right to be Forgotten." Users are the editors of their own lives; Ruva hands them the pen. The project and the demo video are available at http://sisinf00.poliba.it/ruva/.
Paper Structure (8 sections, 3 figures)

This paper contains 8 sections, 3 figures.

Table of Contents

  1. Introduction
  2. The Ruva Architecture
  3. Multimodal Ingestion Workflow
  4. Graph-Grounded Retrieval Workflow
  5. The Deletion Mechanism
  6. Demonstration Scenarios
  7. Evaluation
  8. Conclusion

Figures (3)

  • Figure 1: The Ruva architecture. The Ingestion Workflow transforms multimodal data into semantic triples to populate the Personal Knowledge Graph. The Retrieval Workflow performs graph traversal to generate grounded, hallucination-free answers, entirely on-device.
  • Figure 2: Ruva App Screenshots. \ref{['fig:app_pkg']} shows the inferred PKG, while in \ref{['fig:app_complex_query']}Ruva answers leveraging multimodal data.
  • Figure 3: Judicial scores for Scenario 1 and 2, and $\Delta$ scores for 3.