Resumen
This article presents a method of automated address data extraction from unstructured text on the Internet. The authors focus on the issue of extracting information from the text containing postal addresses and geographical landmarks. The emphasis is on two main techniques: template analysis and statistical analysis with the use of machine learning. The paper describes the advantages of using automated search technologies for Smart Cities and for open data initiatives that are becoming very popular today. In addition, the authors developed software for collecting and retrieving information from the text. The method can be used as a basis for information analysis system on real estate web resources, as well as in semantic web resources and knowledge management systems building.