Temporal JSON

摘要

Web services are the primary suppliers of data on the web, such data is typically formatted using JavaScript Object Notation (JSON) and only the current JSON snapshot is available from a service. But data evolves over time as it is added to, modified, or reduced. Providing an historical view of data is important in many applications. This paper describes how to capture the evolving history of a JSON document and how to support temporal queries. Our approach is to model temporal JSON as a virtual document in which time metadata is mixed with JSON data. The time metadata records when the JSON data is alive. We also describe how to navigate to data within the virtual document. The primary technical contribution of the paper is to show how to represent the virtual model in JSON and how to map temporal path expressions to the representational model. Experiments show that our model is efficient.