Article Preview
TopIntroduction
Today’s society is witnessing a demographic shift towards an increasingly older population (Bloom et al., 2016). Statistical reports (Strulik et al., 2011) show that people aged over sixty represent one-quarter of the population in Europe and that would increase in the next few years. The need to support the elderly in difficulty has prompted researchers to focus on the field of homecare. Several intelligent environmental systems (Luxton et al., 2015) (AI: Ambient Intelligent) have been developed and a significant result in intelligent home technologies, sensor networks, robot assistance and e-health have contributed to the development of the AAL system (Dobre et al., 2016). Today, emerging applications in the field of assistance and support for the elderly are increasingly; these applications are using data from sensors. Multiple AAL-related platforms have been developed, including AMIGO (Georgantas et al., 2005), SOPRANO (Wolf, P., Schmidt et al., 2009), OpenAAL (Wolf et al., 2010), PERSONA (Tazari, Furfari et al., 2010), MPOWER (Stav et al., 2013), UniversAAL (Hanke et al., 2011), HOMER (Fuxreiter et al., 2010), AmiVital (Abril-Jiménez et al., 2009) and OpenCare (Memon et al., 2014). Several factors render the development of AAL system very complex (van den Broek et al., 2010), such as the use of advanced equipment technology like sensors and actuators, the personalization and adaptation of the system to suit various cases, such as its use at home, at work or via mobile support. The localization of elderly people is considered an important task in ambient-assisted living platforms (Stelios et al., 2008). As such, these platforms rely on accurate and secure data to provide the best services suitable to elderly users.
The goal of this paper is to obtain a set of criteria that would be applicable to the context of AAL system, based on ISO/IEC 25012 and ISO/IEC 25010 standards that establish QC on system and software product quality, and data quality, respectively. Given that wireless sensor networks are deployed in many fields, including military operations, mechanical applications, and human services (Shahra et al., 2017), the adopted evaluation procedure will be applied on an indoor positioning platform that continuously determines the positions of people in real-time.
In this work, the definition of data quality model is based on three phases: (i) analyzing and extracting of impact factors, (ii) specifying the data quality criteria, and (iii) the composition of data quality model. Once the data criteria are defined, they are modeled in a hierarchical tree, wherein each factor is composed of one or several criteria and each criterion is composed of one or more sub-criteria until reaching measurable criteria called leaves criteria. This model uses the metrics of the standard models, using equivalence relations, as presented in the paper (Kara, Lamouchi, & Ramdane-Cherif, 2017). To evaluate the entire data quality model for AAL system, an evaluation algorithm was developed in the paper (Kara, Lamouchi, & Ramdane-Cherif, March 2017); in such algorithm, two processes were applied. The first process executes the metrics extracted from the sensors data in order to quantify the leaves criteria; the second process makes use of the result of the first process to evaluate the entire specified data quality model, by using fuzzy logic approach (Klir, & Yuan, 1995) and ends up with a final numerical result.