Information-Retrieval-as-a-Service for the Web of Things: A Survey and a Proposal of IRaaS Architecture

The Cloud age, characterized by a revolution of Data center infrastructures with powerful resources, has leveraged the "as-a-service" paradigms. Nowadays, the Web of Things (WoT) proposes the abstraction of real world entities in a virtual Web avatar, to acquire, process and present real-time information with the ability to connect with the real-world, and to control real things. A new generation of services can arise from these complementary paradigms. One of them is crucial for a daily basis interaction with this new smart world, Information Retrieval (IR), mostly in the form of search engines can also evolve into more powerful tools. Based on that, a new architecture of this kind of services has to be defined by the synergy and challenges that WoT imposes. We propose an IR as a service IRaaS architecture for the new era of the Web of Things. It considers the high dynamics in WoT which leverages a significant amount of changes in the IR collection to merge efficiently conventional IR concepts and the cloud power. Our IRaaS approach takes into account the big-data characteristics of WoT concerning velocity, volume, volatility, and the variety of data. The architecture has two main pillars: the indexing and analysis block, and the querying and retrieval block. We propose building it driven in three dimensions: search scope, resources, and data type. From these perspectives, we show and compare the related works and researches.