This article describes the problem of profiling of objects with varied characteristics. This is relevant for today task, which requires working with large amounts of data and cannot be solved manually. In this paper we describe a solution based on search algorithms of frequent itemsets. The article also presents the results of a comparative analysis of the algorithms (which was the subject of a separate research). The main focus of the article is aimed at substantiation of the choice of an appropriate algorithm based on some characteristic of the source dataset. The study is based on work with actual data of the restaurant industry. Despite this, the results have a wide application, as the approach described in this article can easily be generalized to datasets in any other industry. This article is a logical continuation of the article [5].