Identification of patterns in TV consumption Seven One Media GmbH (EN)

Identification of patterns in TV consumption Seven One Media GmbH (EN)

10:45 - 11:15 l Alternastage l Pattern Recognition & Deep Learning

The prognosis of TV consumption of target groups related to program genres is the planning and steering basis for many continuative applications for a TV station. This includes content topics for programming as well as the seasonal placement of specific programs. In addition the changing media usage due to the rapid increase of internet distribution has also influenced TV consumption. As a TV supplier it’s important to analyze the influence of program-genre-development on TV consumption and station preferences. In the present case a data mining analysis was carried out to answer this specific question and in his presentation Dr. Igor Rotin will show how data mining and analytical methods for prediction, pattern identification and clustering are used in program research. Based on a mathematical cluster model TV consumption per viewer and genre dimension were used as input data. In the clustering process different distance functions were evaluated to determine appropriate similarity measures.

Igor RotinSenior Data Mining Manager, SevenOne Media GmbH (A ProSiebenSat.1 Media SE company)