Skip to main navigation Skip to search Skip to main content

Threshold functional dependencies for time series data

  • Mingyue Ji
  • , Xiukun Wei
  • , Dongjing Miao*
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper extends traditional Functional Dependencies (FDs) to Threshold Functional Dependencies (TFDs) for Time Series Database according to the characteristics of attribute values changing rapidly by time from sensors. In contrast to the unique-to-same pattern in relational schema, TFDs allow determined attribute value within a certain range rather than a clear value when corresponding to the same deciding party. We find that TFDs capable of not only detecting errors resulting from attribute value out-of-bounds in one tuple horizontally, but also from a column of single attribute among several tuples vertically. And we focus more on the former in this article. We draw a clear line between FDs and TFDs because they have some intersection. And we classify TFDs for convenience of research. We provide an inference system for classified TFDs analogous to Armstrong’s axioms, prove its soundness and completeness and explain their differences and connections. We perform some experiments to show effects of TFDs which make some contributions to data quality for Time Series Database.

Original languageEnglish
Title of host publicationDatabase Systems for Advanced Applications. DASFAA 2020 International Workshops - BDMS, SeCoP, BDQM, GDMA, and AIDE, Proceedings
EditorsYunmook Nah, Chulyun Kim, Seon Ho Kim, Yang-Sae Moon, Steven Euijong Whang
PublisherSpringer Science and Business Media Deutschland GmbH
Pages164-174
Number of pages11
ISBN (Print)9783030594121
DOIs
StatePublished - 2020
Event7th International Workshop on Big Data Management and Service, BDMS 2020, 6th International Symposium on Semantic Computing and Personalization, SeCoP 2020, 5th Big Data Quality Management, BDQM 2020, 4th International Workshop on Graph Data Management and Analysis, GDMA 2020, 1st International Workshop on Artificial Intelligence for Data Engineering, AIDE 2020, held in conjunction with the 25th International Conference on Database Systems for Advanced Applications, DASFAA 2020 - Jeju, Korea, Republic of
Duration: 24 Sep 202027 Sep 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12115 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference7th International Workshop on Big Data Management and Service, BDMS 2020, 6th International Symposium on Semantic Computing and Personalization, SeCoP 2020, 5th Big Data Quality Management, BDQM 2020, 4th International Workshop on Graph Data Management and Analysis, GDMA 2020, 1st International Workshop on Artificial Intelligence for Data Engineering, AIDE 2020, held in conjunction with the 25th International Conference on Database Systems for Advanced Applications, DASFAA 2020
Country/TerritoryKorea, Republic of
CityJeju
Period24/09/2027/09/20

Keywords

  • Functional dependency
  • Threshold
  • Time series

Fingerprint

Dive into the research topics of 'Threshold functional dependencies for time series data'. Together they form a unique fingerprint.

Cite this