Web Service Modeling for GraphQL Based College Data Service Access

The Higher Education Database or Pangkalan Data Perguruan Tinggi (PDDIKTI) is a collection of data on the implementation of higher education that is nationally integrated. The data contained in the academic information system of a university must be entered in PDDIKTI. The PDDIKTI Feeder application is one of the services that can be used by every university to help facilitate data input. The large and varied data on the implementation of higher education is an obstacle in the data input process into the PDDIKTI Feeder application. The PDDIKTI Feeder application was developed based on a Web Service with a REpresentational State Transfer (REST) architecture. Web Service using REST Application Programming Interface (API), when a request to an endpoint is executed, it will get additional information that is not really needed. This is because when you access the endpoint, you will get all the data that was determined when the endpoint was developed. So that another filtering stage is needed to separate data that is not needed. The solution to overcome these problems in this research is trying to apply GraphQL. Test scenarios are created by setting up the syntax to access the PDDIKTI Feeder before using GraphQL or using only REST and comparing after implementing GraphQL. The experimental results show that the response time of GraphQL is 20% greater than that of the REpresentational State Transfer (REST). However, the file size response of GraphQL is only 10% compared to REST.