From Sensor Data Streams to Linked Streaming Data: a survey of main approaches

Abstract

Today, large amounts of data are produced by sensor networks. They are continuously producing data streams about real world phenomena. However, these data streams are generated in raw and different formats, lacking the semantics to describe their meanings, which imposes barriers to accessing and using them. To tackle this problem, several solutions using Linked Data Principles have been proposed. In this paper, we survey the main solutions developed by the research communities for publishing data streams in the Web of Data. The major contributions of the paper are the identification of the strengths and limitations of these solutions and, over that basis, the main steps that someone should follow to publish data streams in a manner that anyone can use them, with a minimal understanding of the details. We also highlight the main challenges that emerge from this survey, concluding with a list of research tasks for future work.