SIMIT: Subjectively Interesting Motifs in Time Series
Numerical time series data are pervasive, originating from sources as diverse as wearable devices, medical equipment, to sensors in industrial plants. In many cases, time series contain interesting information in terms of subsequences that recur in approximate form, so-called <i>motifs</i&g...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2019-06-01
|
Series: | Entropy |
Subjects: | |
Online Access: | https://www.mdpi.com/1099-4300/21/6/566 |
_version_ | 1811277944374427648 |
---|---|
author | Junning Deng Jefrey Lijffijt Bo Kang Tijl De Bie |
author_facet | Junning Deng Jefrey Lijffijt Bo Kang Tijl De Bie |
author_sort | Junning Deng |
collection | DOAJ |
description | Numerical time series data are pervasive, originating from sources as diverse as wearable devices, medical equipment, to sensors in industrial plants. In many cases, time series contain interesting information in terms of subsequences that recur in approximate form, so-called <i>motifs</i>. Major open challenges in this area include how one can formalize the interestingness of such motifs and how the most interesting ones can be found. We introduce a novel approach that tackles these issues. We formalize the notion of such subsequence patterns in an intuitive manner and present an information-theoretic approach for quantifying their interestingness with respect to any prior expectation a user may have about the time series. The resulting interestingness measure is thus a <i>subjective</i> measure, enabling a user to find motifs that are truly interesting <i>to them</i>. Although finding the best motif appears computationally intractable, we develop relaxations and a branch-and-bound approach implemented in a constraint programming solver. As shown in experiments on synthetic data and two real-world datasets, this enables us to mine interesting patterns in small or mid-sized time series. |
first_indexed | 2024-04-13T00:25:54Z |
format | Article |
id | doaj.art-a10a5ebbed394462aa51a80036f0c7b7 |
institution | Directory Open Access Journal |
issn | 1099-4300 |
language | English |
last_indexed | 2024-04-13T00:25:54Z |
publishDate | 2019-06-01 |
publisher | MDPI AG |
record_format | Article |
series | Entropy |
spelling | doaj.art-a10a5ebbed394462aa51a80036f0c7b72022-12-22T03:10:36ZengMDPI AGEntropy1099-43002019-06-0121656610.3390/e21060566e21060566SIMIT: Subjectively Interesting Motifs in Time SeriesJunning Deng0Jefrey Lijffijt1Bo Kang2Tijl De Bie3Department of Electronics and Information Systems, Ghent University, Technologiepark-Zwijnaarde 122, 9052 Ghent, BelgiumDepartment of Electronics and Information Systems, Ghent University, Technologiepark-Zwijnaarde 122, 9052 Ghent, BelgiumDepartment of Electronics and Information Systems, Ghent University, Technologiepark-Zwijnaarde 122, 9052 Ghent, BelgiumDepartment of Electronics and Information Systems, Ghent University, Technologiepark-Zwijnaarde 122, 9052 Ghent, BelgiumNumerical time series data are pervasive, originating from sources as diverse as wearable devices, medical equipment, to sensors in industrial plants. In many cases, time series contain interesting information in terms of subsequences that recur in approximate form, so-called <i>motifs</i>. Major open challenges in this area include how one can formalize the interestingness of such motifs and how the most interesting ones can be found. We introduce a novel approach that tackles these issues. We formalize the notion of such subsequence patterns in an intuitive manner and present an information-theoretic approach for quantifying their interestingness with respect to any prior expectation a user may have about the time series. The resulting interestingness measure is thus a <i>subjective</i> measure, enabling a user to find motifs that are truly interesting <i>to them</i>. Although finding the best motif appears computationally intractable, we develop relaxations and a branch-and-bound approach implemented in a constraint programming solver. As shown in experiments on synthetic data and two real-world datasets, this enables us to mine interesting patterns in small or mid-sized time series.https://www.mdpi.com/1099-4300/21/6/566time seriesmotif detectioninformation theorysubjective interestingnesspattern miningexploratory data mining |
spellingShingle | Junning Deng Jefrey Lijffijt Bo Kang Tijl De Bie SIMIT: Subjectively Interesting Motifs in Time Series Entropy time series motif detection information theory subjective interestingness pattern mining exploratory data mining |
title | SIMIT: Subjectively Interesting Motifs in Time Series |
title_full | SIMIT: Subjectively Interesting Motifs in Time Series |
title_fullStr | SIMIT: Subjectively Interesting Motifs in Time Series |
title_full_unstemmed | SIMIT: Subjectively Interesting Motifs in Time Series |
title_short | SIMIT: Subjectively Interesting Motifs in Time Series |
title_sort | simit subjectively interesting motifs in time series |
topic | time series motif detection information theory subjective interestingness pattern mining exploratory data mining |
url | https://www.mdpi.com/1099-4300/21/6/566 |
work_keys_str_mv | AT junningdeng simitsubjectivelyinterestingmotifsintimeseries AT jefreylijffijt simitsubjectivelyinterestingmotifsintimeseries AT bokang simitsubjectivelyinterestingmotifsintimeseries AT tijldebie simitsubjectivelyinterestingmotifsintimeseries |