Implementation of the Concept of a Repository for Automated Processing of Semi-Structural Data

Semi-structural data tend to be problematic due to the sparsity of their attributes and due to the fact that, regardless of their type, they are immensely diverse. This means that data storage is a challenge, especially when the data contained within a relational database – often a strict requirement defined in advance. In this paper, we present a thoroughly described concept of a repository that is capable of storing and processing semi-structural data. Based on this concept, we establish a database model comprising the architecture and the tools needed to search the data and build relevant processors. The processor described may assign roles and dispatch tasks between the users. We demonstrate how the capacities of this repository are capable of overcoming current limitations by creating a system for facilitated digitization of scientific resources. In addition, we show that the repository in question is suitable for general use, and, as such, may be adapted to any domains in which semi-structural data are processed, without any additional work required. Keywords—document management system, ECM, JSON, workflow.