User segmentation in online advertising using latent Dirichlet allocation

Darius Aliulis, Vytautas Janilionis
2015 Lietuvos matematikos rinkinys  
User segmentation is one the most important problems in online advertiting. The use of online latent Dirichlet allocation model for analysing big datasets for this purpose is proposed in this paper. The relationship between the number of segments and segmentation quality metrics is analized and criteria for assigning ads to user segments are proposed.
doi:10.15388/lmr.b.2015.01 fatcat:gixqdyurw5cyraiyp5l4m7xr6e