This paper presents a supervised classification model, where the indicators of correlation between dependent and independent variables within each class are utilized for a transformation of the large-scale input data to a lower dimension without loss of recognition relevant information. In the case study, we use the consumption data recorded by smart electricity meters of 4200 Irish dwellings along with half-hourly outdoor temperature to derive 12 household properties (such as type of heating, floor area, age of house, number of inhabitants, etc.). Survey data containing characteristics of 3500 households enables algorithm training. The results show that the presented model outperforms ordinary classifiers with regard to the accuracy and temporal characteristics. The model allows incorporating any kind of data affecting energy consumption time series, or in a more general case, the data affecting class-dependent variable, while minimizing the risk of the curse of dimensionality. The gained information on household characteristics renders targeted energy-efficiency measures of utility companies and public bodies possible.
Energy utility companies hold a great potential to improve sustainable practices, as they have access to large amounts of energy consumption data that can be translated into actionable insights toward individual consumers. We investigate the role of energy informatics based artifacts for utility companies to (i) profile households and to individually address customers that have a large potential to elevate their energy efficiency and (ii) at the same time create revenue by purchasing energy efficient goods and related services. Furthermore, we demonstrate how the predicted attributes of millions of households can, in turn, contribute to the economic, environmental, and societal value creation for energy suppliers, utility customers, and the community as a whole. For our empirical study on the detection of households with old heating systems, we cooperated with a large energy supplier in Belgium. The results are used for targeted positioning of new heating systems in a cross-selling campaign and may serve as a blueprint for related data analytics artifacts in this field.
Utility companies generally have an extensive customer base, yet their knowledge about individual households is small. This adversely affects both the development of innovative, household specific services and the utilities’ key performance indicators such as customer loyalty and profitability. With the goal to overcome this knowledge deficit, persuasive systems in the form of customer self-service applications and efficiency coaching portals are becoming the getaway of data exchange between utility and user. While improved customer interaction and the collection of customer data within respective information systems is an important step towards a service-oriented company, the immediate value generated from the collected data is still limited, mostly due to the small fraction of customers actually using such systems. We show how to utilize the knowledge gained from the sparse number of active web users in order to provide low-cost and large-scale insights to potentially all residential utility customers. We do so using machine-learning-based Green IT artifacts that allow for improving decision-making, effectiveness of energy audits, and conservation campaigns, thus ultimately increasing the customer value and adoption of related services. Moreover, we show that data from the publically available geographic information systems can considerably improve the decision quality.
Within the last decade, several studies have analysed the life writing produced by veterans of the U.S. American war against and the ensuing occupation of Iraq (2003-2011). These studies are beginning to offer insightful analyses of several autobiographies written by Iraq War veterans, they focus almost exclusively on male combat veterans, perpetuating a bias that has shaped the genre of autobiographies of war in the United States and the subsequent scholarly analyses of these autobiographies for decades, if not centuries. My paper begins to address this bias by focusing on two autobiographies by female soldiers, namely Jane Blair’s Hesitation kills (2011) and Heidi Squier Kraft’s Rule number two (2007). As the nineteenth-century biographies that Jo Burr Margadant has analysed, these texts can be considered ‘collage[s] of familiar notions merged in unfamiliar ways’ (Margadant, 2000, p.2), as both of them portray the experiences of female soldiers in a combat zone and thus trouble those assumptions about combat, identity, and gender that continue to structure more traditional American life writing about war. In my analysis I show how the autobiographical subjects in these texts are constructed through what I call ‘military femininities’. Drawing in part on conceptual and theoretical work done by Judith Butler, Leigh Gilmore, and R.W. Connell I argue that these military femininities are instrumental in creating female autobiographical subjects that are authorized to talk about war in the cultural context of the contemporary United States, and thus in establishing soldierly identities for American women more generally.