Ontology-based feature generation to improve accuracy of activity recognition in smart environments. (May 2018)

• Record Type: Journal Article
• Authors: Salguero, A.G.; Espinilla, M.
• Is Part Of: Computers & electrical engineering. Volume 68(2018)
• Page Start: 1
• Page End: 13
• Publication Date: 2018-05
• Subjects: Activity recognition -- Smart environments -- Ambient assisted living (AAL) -- Activities of Daily Living (ADL) -- Ontology -- Data-Driven approaches -- Knowledge-Driven approaches; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2018.03.048
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: In recent years, many techniques have been proposed for automatic recognition of Activities of Daily Living from smart home sensor data. However, classifiers usually use features created ad hoc. In this work, the use of ontologies is proposed for the fully automatic generation of these features. The process consists of converting the original dataset into an ontology and then combine all the concepts and relations in that ontology to obtain relevant class expressions. The high formalization of ontologies allows us to reduce the search space by discarding many meaningless expressions, such as contradictory or unsatisfiable expressions. The relevant class expressions are then used as features by the classifiers to build the classification model. To validate our proposal, we have used as reference the results obtained by four different classification algorithms that use the most commonly used features.