Energy efficiency and sustainability are important factors to address in the context of smart cities. In this sense, smart metering and nonintrusive load monitoring play a crucial role in fighting energy thefts and for optimizing the energy consumption of the home, building, city, and so forth. The estimated number of smart meters will exceed 800 million by 2020. By providing near real-time data about power consumption, smart meters can be used to analyze electricity usage trends and to point out anomalies guaranteeing companies' safety and avoiding energy wastes. In literature, there are many proposals approaching the problem of anomaly detection. Most of them are limited because they lack context and time awareness and the false positive rate is affected by the change in consumer habits. This research work focuses on the need to define anomaly detection method capable of facing the concept drift, for instance, family structure changes; a house becomes a second residence, and so forth. The proposed methodology adopts long short term memory network in order to profile and forecast the consumers' behavior based on their recent past consumptions. The continuous monitoring of the consumption prediction errors allows us to distinguish between possible anomalies and changes (drifts) in normal behavior that correspond to different error motifs. The experimental results demonstrate the suitability of the proposed framework by pointing out an anomaly in a near real-time after a training period of one week. INDEX TERMS Anomaly detection, concept drift, machine learning, smart grid, time series analysis. I. INTRODUCTION A. CONTEXT
Semantic annotation is at the core of Semantic Web technology: it bridges the gap between legacy non-semantic web resource descriptions and their elicited, formally specified conceptualization, converting syntactic structures into knowledge structures, i.e., ontologies. Most existing approaches and tools are designed to deal with manual or semi-/automatic semantic annotation that exploits available ontologies through the pattern-based discovery of concepts. This work aims to generate the automatic semantic annotation of web resources, without any prefixed ontological support. The novelty of our approach is that, starting from web resources, content with a high-level of abstraction is obtained: concepts, connections between concepts, and instance-population are identified and arranged into an ex-novo ontology. The framework is designed to process resources from different sources (textual information, images, etc.) and generate an ontology-based annotation. A data-driven analysis reveals the data and their intrinsic relationships (in the form of triples) extracted from the resource content. On the basis of the discovered semantics, corresponding concepts and properties are modeled, allowing an ad hoc ontology to be built through an OWL-based coding annotation. The benefit of this approach is the generation of knowledge structured in a quite automatic way (i.e., the human support is restricted to the configuration of some parameters). The approach exploits a fuzzy extension of the mathematical modeling of Formal Concept Analysis and Relational Concept Analysis to generate the ontological structure of data resources
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.