Clustering of Physical Behaviour Profiles using Knowledge-intensive Similarity Measures

Deepika Verma, Kerstin Bach, Paul Mork

Abstract

In this paper, we reuse the Case-Based Reasoning model presented in our last work (Verma et al., 2018) to create a new knowledge intensive similarity-based clustering method that clusters a case base such that the intra-cluster similarity is maximized. In some domains such as recommender systems, the most similar case may not always be the desired one as a user would like to find the closest, yet significantly different cases. To increase the variety of returned cases, clustering a case base first, before the retrieval is executed increases the diversity of solutions. In this work we demonstrate a methodology to optimize the cluster coherence as well to determine the optimal number of clusters for a given case base. Finally, we present an evaluation of our clustering approach by comparing the results of the quality of clusters obtained using our knowledge intensive similarity-based clustering approach against that of the state-of-the-art K-Means clustering method.

Download


Paper Citation


in Harvard Style

Verma D., Bach K. and Mork P. (2020). Clustering of Physical Behaviour Profiles using Knowledge-intensive Similarity Measures.In Proceedings of the 12th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART, ISBN 978-989-758-395-7, pages 660-667. DOI: 10.5220/0008980406600667


in Bibtex Style

@conference{icaart20,
author={Deepika Verma and Kerstin Bach and Paul Mork},
title={Clustering of Physical Behaviour Profiles using Knowledge-intensive Similarity Measures},
booktitle={Proceedings of the 12th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART,},
year={2020},
pages={660-667},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0008980406600667},
isbn={978-989-758-395-7},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 12th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART,
TI - Clustering of Physical Behaviour Profiles using Knowledge-intensive Similarity Measures
SN - 978-989-758-395-7
AU - Verma D.
AU - Bach K.
AU - Mork P.
PY - 2020
SP - 660
EP - 667
DO - 10.5220/0008980406600667