WissKI is subject to the possibilities and limitations of its programming and technical equipment; but beside that, its data model significantly influences its informative value and usability. Data modeling is a highly complex task, especially when you are dealing with imprecision, uncertainty, ambiguity, or missing data (as is especially common in the digital humanities.)
Before you start modeling, you should have a good overview of the data mass (how many entities), data density (how many properties per entity), and be aware of the specifics and gaps of the material and the requirements and functionalities of the research environment.
It is strongly recommended to create a meaningful documentation from the beginning of your project and to visualize the data model (software like Visio or Lucidchart can help).