Binary data in matrices with singular value decomposition method

A big challenge that faces many applications in different fields suffers in dealing with datasets of massive size. Additionally, retrieving and casting this data is somewhat time-consuming. Applications such as government or any institution election, surveys, healthcare …etc., leverage techniques of data reduction, dimensionality reduction, matrix decomposition, or compression such as the Singular Value Decomposition Technique. Our paper shows the use of this technique as a method in certain circumstances where data is of binary type and can be retrieved, cast, or updated in less time and in a smaller size without losing any information. In other words, we prove practically that the massive size of binary values can be managed in a form of matrices with low rank (low rank is one of the bases used in the Singular Value Decomposition technique) to return the exact matrix of information instead of dealing with the original large matrix of data. The experimental results are implemented on a Lenovo machine, Intel Corei5, CPU 2.5GH with 8GB of RAM, using visual basic, C#, in Visual Studio 2019 environment.