Abstract

During past years, availability of new or alternative data sources showed wide range of possible applications in different official statistics topics. In price statistics, scanner data from private data owners is of high importance for future improvements of HICP calculations. Nevertheless, the application or integration of these kind of data into production process requires detailed analysis and preparation work. In the presentation we will give a short overview of process integrating scanner data, automated quality checks, classification process. We will present some results on application of machine learning methods for classification of data.