Posts by Collection

Towards User-Centric Graph Repairs

IEEE 40th International Conference on Data Engineering Workshops (ICDEW), 2024

An early vision and problem formulation for user-centric, interactive repair of inconsistent graph data, combining automatic inference with human expertise.

Recommended citation: A. Pachera, A. Bonifati and A. Mauri, "Towards User-Centric Graph Repairs," in 2024 IEEE 40th International Conference on Data Engineering Workshops (ICDEW), Utrecht, Netherlands, 2024, pp. 375-376, doi: 10.1109/ICDEW61823.2024.00057.
Download Paper

Understanding Students’ Errors in Graph Query Formulation

ACM Transactions on Computing Education (ToCE), 2025

An empirical study of the common errors students make when formulating graph queries, providing insights for the design of graph database education tools.

Recommended citation: Understanding Students' Errors in Graph Query Formulation. ACM Transactions on Computing Education (ToCE), 2025. DOI: 10.1145/3743687
Download Paper

Grafixer: Enabling User-Centric Repairs for Property Graphs

PACMMOD — Demonstration, SIGMOD 2025

An interactive tool for human-in-the-loop repair of property graphs. Users upload datasets, define constraints in Cypher, and collaboratively correct inconsistencies; administrators monitor real-time statistics and oversee the repair process through an interactive dashboard.

Recommended citation: A. Pachera, A. Bonifati, A. Mauri. Grafixer: Enabling User-Centric Repairs for Property Graphs. PACMMOD Demonstration, SIGMOD 2025. DOI: 10.1145/3722212.3725105
Download Paper

User-Centric Property Graph Repairs

Proceedings of the ACM on Management of Data (PACMMOD), SIGMOD 2025

An interactive, user-centric approach to repairing property graphs under denial constraints. It introduces a query-based inconsistency detection mechanism, a dependency graph for tracking violations, and an assignment algorithm enabling multi-user repairs via independent sets — outperforming interactive and non-interactive baselines by 30% on average, validated with a user study.

Recommended citation: A. Pachera, A. Bonifati, A. Mauri. User-Centric Property Graph Repairs. Proceedings of the ACM on Management of Data (PACMMOD), 2025. DOI: 10.1145/3709735
Download Paper

WhatIf: Causal Analysis with Graph Databases

Proceedings of the VLDB Endowment (PVLDB), 2025

We propose graph-based techniques for interactive causal analysis over property graph databases, enabling efficient exploration of causal relationships in complex, interconnected data.

Recommended citation: A. Pachera, A. Bonifati, A. Mauri. WhatIf: Causal Analysis with Graph Databases. Proceedings of the VLDB Endowment (PVLDB), 2025. DOI: 10.14778/3749646.3749671
Download Paper

Budgeted Interactive Property Graph Repair with Graph Neural Networks

ACM SIGMOD 2027 (To Appear)

A Graph Neural Network–based framework for interactive property graph repair that combines representation learning, optimization, and human expertise under budget constraints. The model learns to estimate repair difficulty and decides, within a fixed budget, which repairs to automate and which to route to human experts.

Recommended citation: A. Pachera, A. Bonifati, A. Mauri. Budgeted Interactive Property Graph Repair with Graph Neural Networks. ACM SIGMOD 2027 (to appear).
Download Paper

Improving Data Quality and Reasoning in Graph Databases: Human-in-the-Loop, Machine Learning, and Causal Approaches

Published: March 12, 2026

Research seminar given during my visiting period at the University of British Columbia, hosted by Prof. Laks V. S. Lakshmanan. The talk surveyed my work on improving data quality and reasoning in graph databases, spanning human-in-the-loop repair, graph neural networks, and causal analysis over property graphs.

Bases de données avancées [FR]

Workshop in Undergraduate Program · Université Claude Bernard Lyon 1, Math Department · 2024

Github du cours

Laboratory of Data Processing and Analytics

Workshop in Master Program · Université Claude Bernard Lyon 1, Math Department · 2024

The objective of this course is to introduce students to the principles and methods of advanced data management and processing. The course covered the techniques of storing, pre-processing and processing different types of data (structured, semi-structured and unstructured).

Laboratory of Soft Skill 1

Workshop in Master Program · Université Claude Bernard Lyon 1, Math Department · 2024

Available projects :

Foundations of Prompt Engineering

Lecturer · Bocconi University · 2025

Invited lecturer for a module on the Foundations of Prompt Engineering, introducing students to large language models and effective prompting strategies for real-world tasks.