AeL

Data analysis Tool

Web oriented solution for data analysis for results from learning courses including work with huge amount of data, data surveys etc.

It is an integrated solution for data gathering from different sources. All data are stored in SQL standardised DB with open XML structures. Bring "one click" solution for data analysis over data from learners including graphical interpretation, calculation different forms of correlations, intercorrelations etc. Unique detection system for finding unknown relations between results. System of computation of correlations between learners results.

Data analysis Tool


It is a system with a standard division into several parts:

  • Part 1: Front-end The Front-end part made in programming language PHP version 7 using HTML, CSS and JavaScript functionality. This approach allows for easy portability and sufficient functionality in relation to analytical functions. It has also been shown that, in spite of the initial worries about a potentially lower performance for analytical calculations, this solution, despite the large volume of data, is fully functional. Another important aspect is the possibility of integrating communication with other software modules and external software such as the LMS system. In this case, the system is equipped with a web services interface that is industrial standard.
  • Part 2: Back-endThe Back-end part was made in C# and running as a service that performs calculations that simplify and speed up the display. A typical example is the pre-counting of correlation tables. Another case is the processing of scanned questionnaires (OCR functions).
  • Part 3: Database As a database server, a MySQL database has been chosen, which is otherwise sufficient for performance purposes.
  • The last, internal part is then the part of the system that is intended for the mathematical/analytical model. As the data contained in the database turned out to be quite suitable for modelling. Data can be divided into input (before the course) and output (after the course). Subsequently, based on the system theory, it is possible to identify the model of relational relation between input-output. In this case, the variability of the various models and the interpretation of the data, including, for example, extrapolation for cases not contained in the database, are extremely inspiring.

Technical solution brings many advantages. One of the most important are the very wide possibilities of filtering and formatting data before processing it for analytics. This is due to the use of very complex SQL queries that allow data to be interpreted over the entire database structure. Therefore, it is easier to separate the data preparation time before it is processed, which in principle is no longer possible in Microsoft Excel office applications.

One example we can demonstrate follow.

Correlation graph


Z axis is correlation coef. For questions results (X and Y axes). Higher peak means better correlation.

One of the advantages of a good database is the possibility of qualitative and quantitative work with the graphical output. To take advantage of this option, we have calculated a correlation coefficient according to Pearson for all questions in the system (but of course, a different approach, including T-test, etc.). These results can then be displayed in a 3D graph showing the correlation between the questions in a clear and clear way - a higher value towards 1 means a greater correlation.

Contact to local distributor Email: IE CZ SK EN RO
Tel.: +353 1 700 7648
Contact to global Newtelp staff Email: gabriel.muntean@dcu.ie
Tel.: +353 1 700 7648