Thesis Abstract: Interactive Subgroup Discovery for the Conversational Data Governance Platform "Talking to your Data"

Astrid Sierens, Isel Grau, Luis Daniel Hernandez, Simeon Michel, Vicky Froyen, Catherine Middag, Ann Nowé

Research output: Chapter in Book/Report/Conference proceedingConference paper

Abstract

In this master thesis, a new interactive subgroup discovery algorithm is proposed. This method has two main contributions. First, the algorithm allows the expert to intervene during the search process by assessing each subgroup with a degree of appreciation which influences the search process. The second contribution is a diversity parameter that allows the user to avoid that the new subgroups share more than a chosen percentage of instances with already found subgroups. Experiments show that when diversity control is performed, the resulting subgroups have less overlap than the baseline version of the algorithm. Additionally, when using the proposed interactive version of the algorithm, a higher user appreciation of the subgroups is observed. This interactive subgroup discovery algorithm was implemented in the backend of a conversational agent for supporting business analysts in data mining tasks.
Original languageEnglish
Title of host publicationProceedings of the 33rd Benelux Conference on Artificial Intelligence and the 30th Belgian-Dutch Conference on Machine Learning BNAIC/BeneLearn 2021
Pages769-771
Number of pages3
Publication statusPublished - 2021
Event33rd Benelux Conference on Artificial Intelligence and 30th Belgian-Dutch Conference on Machine Learning - , Luxembourg
Duration: 10 Nov 202112 Nov 2021
https://bnaic2021.uni.lu/

Conference

Conference33rd Benelux Conference on Artificial Intelligence and 30th Belgian-Dutch Conference on Machine Learning
Abbreviated titleBNAIC/BeneLearn 2021
CountryLuxembourg
Period10/11/2112/11/21
Internet address

Fingerprint

Dive into the research topics of 'Thesis Abstract: Interactive Subgroup Discovery for the Conversational Data Governance Platform "Talking to your Data"'. Together they form a unique fingerprint.

Cite this