Mining for Patterns of Behaviour in Children with Autism Through Smartphone Technology

William Burns, Mark Donnelly, Nichola Booth

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

2 Citations (Scopus)

Abstract

A requirement to maintain detailed recording of child behaviour is commonplace for families engaged in home-based autism intervention therapy. Periodically, a Behaviour Analyst reviews this data to formulate new behaviour change plans and as such, the quality and accuracy of data is paramount. We present a smartphone application that aims to streamline the traditional paper based approaches, which are prone to non-compliance and erroneous detail. In addition, we have applied association rule mining to the collected behaviour data to extract patterns in terms of behaviour causes and effects with a view to offer intelligent support to the Behaviour Analysts when formulating new interventions. The paper outlines the results of a small evaluation of the smartphone component before introducing the methodology used to mine that data to highlight behaviour rules and patterns. Consequently, based on an initial sample of child behaviours, the methodology is then compared to a Behaviour Analyst’s assessment of corresponding paper based records.
Original languageEnglish
Title of host publicationUnknown Host Publication
PublisherSpringer
Pages147-154
Number of pages7
Volume8456
DOIs
Publication statusPublished - Jun 2014
Event12th International Conference on Smart Homes and Health Telematics, ICOST 2014; - Denver; United States
Duration: 1 Jun 2014 → …

Conference

Conference12th International Conference on Smart Homes and Health Telematics, ICOST 2014;
Period1/06/14 → …

Keywords

  • Association rule mining
  • Autism spectrum disorders
  • Behaviour monitoring
  • Health records
  • Intelligent data analysis
  • Smartphone

Fingerprint Dive into the research topics of 'Mining for Patterns of Behaviour in Children with Autism Through Smartphone Technology'. Together they form a unique fingerprint.

Cite this