Linked data platform for web applications

Abstract

Most of today's web applications are tightly coupled to proprietary server backends that store and control all user data. This thesis presents Linked Data as a decentralized web app platform, eliminating vendor lock-in, and separating user data from web apps giving users control over their data and where it's stored, independent of choice of application. Linked Data architecture replaces traditional app-data silos with a universal integration platform using global identifiers, shared ontologies, and a scalable, standardized data model. We provide 3 interoperable Linked Data server implementations in PHP, Python, and Go, and evaluate their performance. Traditional filesystems and relational databases have been integrated, and several new decentralized web apps have been developed for the platform. As of May 2014, worldwide open source community members are using and contributing to these compatible apps and servers, and the design continues to be refined and standardized at the W3C. All code available at: https://github.com/linkeddata