The National Bank of Georgia continues to promote data science. The Bank was pleased to host data science enthusiasts November 5-15 who were invited to test their skills on Texprint.

 

The purpose of the Texprint was to create a model for converting high-precision image to text (OCR) which would make it possible to read Georgian documents of different formats (font, quality, mixed with other languages) with high accuracy and convert the given record into digital form. Special attention was paid to the extraction of identification data from the ID card.

 

In identifying winners, preference was given to the adequacy of the approach taken to solve the problem and the accuracy of the prediction. Considering these two components, the team with Soso Sukhitashvili and Orhan Shamilov showed the best results.

 

The NBG is eager to continue promoting data science in the form of various projects and competitions.