From representational state transfer (REST) to representational action state transfer (REAST): enabling smarter web interactions for web tasking

  • Authors:
  • Joanna W. Ng

  • Affiliations:
  • IBM Canada Software Laboratory, CAS Research

  • Venue:
  • CASCON '13 Proceedings of the 2013 Conference of the Center for Advanced Studies on Collaborative Research
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Today, the web is designed for information retrieval. Its current interaction model is web browsing. However, browsing is not an intuitive interaction model for web tasking. This paper provides a solution by introducing an architectural style called REpresentational Action State Transfer (REAST), extended from the classic architectural style of REpresentational State Transfer (REST). REAST represents the action state of the resource instead of the resource data state as in REST. By transferring the representation of action state of the resource between the requesting processer and the resource server in the architectural style of REAST, instead of transferring the data state of the resource as in the architectural style of REST, machine operated web interfaces with the dispensability of human web users is, for the first time, made possible. This opens up opportunities for new interaction model design that are better fit for web tasking where web browsing falls short. With its RESTful inheritance, REASTful resources share universal resource representation and the unified interface, thus have the architecturally built-in interoperability between enterprise resource items and items from the Internet of Things, in both machine automation and mixed initiatives between machine and human web users. The RESTful inheritance also means that REASTful resources relate in execution under the constraint of hypermedia controls. This enables the building of resource oriented intelligent web agents that can work generically across the web without domain specificity and can execute on actions that are non-preprogrammed. This paper coined such Resource Oriented Intelligent Agents 'the BOT'.