QUARRY: A Graph Model for Queryable Association Rules

Research output: Chapter in Book/Conference paperConference paperpeer-review


Association rule mining is a pivotal technique for knowledge discovery, but often involves time-intensive manual labour when performed on large datasets. In this paper we propose a solution for this problem: QUARRY, a graph model that enables consumable and queryable insights from association rules. In contrast to existing systems which take a list of rules and display them in a purpose-built visualisation, our graph-based model enables association rules to be queried directly via graph queries. Through a case study on maintenance data we show how this model enhances knowledge discovery by eliminating the need for domain experts to trawl through large lists of rules to find useful information. QUARRY, which is designed for compatibility with existing knowledge graphs, provides users with the means to easily search for rules pertaining to specific items as well as roll up and drill down on their searches using the concept hierarchy. Domain experts may also query for association rules based on transaction properties such as costs and dates, enabling critical insights into their data.
Original languageEnglish
Title of host publicationAI 2022
Subtitle of host publicationAdvances in Artificial Intelligence - 35th Australasian Joint Conference, AI 2022, Proceedings
EditorsHaris Aziz, Débora Corrêa, Tim French
PublisherSpringer Science + Business Media
Number of pages14
ISBN (Print)9783031226946
Publication statusPublished - 2022
Event35th Australasian Joint Conference on Artificial Intelligence, AI 2022 - Perth, Australia
Duration: 5 Dec 20229 Dec 2022

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13728 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference35th Australasian Joint Conference on Artificial Intelligence, AI 2022


Dive into the research topics of 'QUARRY: A Graph Model for Queryable Association Rules'. Together they form a unique fingerprint.

Cite this