Information Retrieval System Based on the Knowledge Extraction

The article discusses stages in developing a method for solving the search problems in the natural language processing. It reveals relevance of the topic. The article proposes a method for optimizing the search for the most relevant documents using the semantic knowledge graph in the context of modern approaches to the natural language processing. It describes the main stages aimed at improving the accuracy in searching for documents that match a complex user query. Results of the information retrieval algorithm operation are analyzed, and strengths and weaknesses of the approach are highlighted. The article provides theoretical computation of increasing the quality. Efficient introduction of the proposed method could improve the quality of searching for the composite entities.