You are here

The system architecture and technical details

21 January, 2016 - 14:46

The architecture of the system is available in (Figure : General Architecture of the Project Management E-assessment System).

The system has several components, which ensures its flexibility:

  • the admin module (offers the possibility to accomplish operations on levels, competences, trainers, users or questions);
  • the trainer module (allows the creation of rules-based tests, visualization of previously created tests and visualization of students’ information);
  • the student module (contains the web application used by the students to resolve online tests, which are created with various adaptation models);
  • the web service;
  • the business services, which use a domain model, created with nHibernate technology;
  • the database, which contains information about questions, tests, simulations, users;
  • the database is relational and is constructed using the IPMA Competences Baseline (International Project Management Association, 2006).
media/image7.png
Figure 8.6 General Architecture of the Project Management E-assessment System 
media/image8.png
Figure 8.7 Excerpt from the Class Diagram of the Business Services 

The communication with the database is made in a formal manner:

  • the admin and the trainer module communicate with the database via a set of business services (an excerpt from the class diagram of the business services is available Figure 8.7);
  • the student module communicates with the database via the web service, which receives Ajax calls (from Dojo library) (see Figure 8.8).
media/image9.png
Figure 8.8 Web Service Architecture 
media/image10.png
Figure 8.9 The Class Diagram of the PSO Algorithm 

The solution was developed in C#, the exception being made by the web student module, which was developed in Javascript. The calls to the web service were implemented using the Dojo library. The domain model was created with nHibernate. The formative character of the e-assessment system is given by the adaptive test building engine and the feedback module, from Figure 8.8. The feedback engine offers the trainees the possibility of revising their mistakes and consulting further web bibliography recommendations, to fill their knowledge gaps. Thus, future learning directions are offered. The adaptive engine contains the PSO algorithm applied to test construction and a rules-based algorithm, where a set of rules for building a test are established via the trainer module. The classes diagram used by the PSO algorithm is shown in Figure 8.9. Thanks to the use of abstract classes (see class PSO and Particle), the minimizing function can be easily changed. All the parameters which have to be set as prerequisites are fields in FitnessFunction (presented theoretically in formula 1): they are seen as constants by the FitnessFunction.