The paper deals with implementation of Descriptive Similarity for numeric estimation of the similarity of the elective programs of political parties. The method - the assessment of descriptive similarity - based on simple arithmetic operations and the use of so-called equalization’s tables is easy to use for academic political scientists and policy observers. Method of estimation statements’ similarity is very crucial for scientific areas such as management, decision-making process, military, politics. The programs of nine political parties were analysed. The authors studied the parties from the big number of parties. The amazing comparisons were received during investigation. We could see that for politics it can be useful in order to prove the ideology of a political party and check the populistic statements. The proposed approach to the automated estimation of the similarity coefficient is important for the formation of applied artificial intelligence. This will allow not only to quickly assess the similarity of programs but also to form a basis for machine learning to confirm the existing ideology of the party or its absence.