At our client from the sports industry, we introduced a robot that uses a machine learning algorithm in the KNIME application. The robot's working pattern is to redirect requests from the queuing system to the appropriate group for the request. Assigning the request to the appropriate group is done in two steps:
01. Defined keywords with their assigned queue names are checked. Then, after verification with the title of such a request, it is redirected to the selected queue.
02. In case of missing keywords from the first method, the robot with the help of a machine learning algorithm implemented through the KNIME app tries again to assign a correct queue to the request based on the title.
The robot's final step is to redirect the request to the queue selected in point 1 or 2.
In addition, every two months the machine learning algorithm is re-trained using tickets from the last six months so that it is always up to date.