In recent years, interest in knowledge graphs (KG) has increased exponentially in both the scientific and industrial communities. KGs play an important role in AI applications such as natural language processing including question-answering systems, recommender systems, and search engines. Integration of different KGs is one of the most pressing problems and is used, for example, to develop complex digital twins of industrial systems. One of the components of the KG integration problem is the entity alignment problem, which attempts to identify entities in different KGs that describe the same real-world object. A special case of this problem is the problem of cross-language entity identification, which is closely related to the problem of import substitution, such as finding equivalent drugs, spare parts, or devices for the Internet of Things. Unfortunately, in real KGs, many entities may not have equivalents in another KG. This paper describes entity alignment experiments using the example of a Russian-English dataset with unmatchable entities.

knowledge graph, entity alignment, unmatchable entities