Introduction. The paper is devoted to the problem of studying the semantic space of texts on physics and astronomy using computer technology. The paper purpose is to: 1) determining the semantic distances between the key concepts of the microcosm in physical and astronomical educational texts; 2) constructing graphs visualizing the semantic spaces of these texts and comparing them. Materials and methods. The analysis used texts on physics and astronomy from school and university textbooks, as well as Wikipedia articles (total volume is 100 thousand words). Text processing and calculation of the cosine measure of concept proximity is carried out using computer programs written in ABCPascal and Excel spreadsheets; an online Internet resource is used to build graphs. Research results. Two files with texts on physics and astronomy are obtained, key concepts of microcosm physics are identified, and cosine proximity measures and semantic distances are determined for them. Graphs are constructed; their vertices correspond to concepts, and the edges characterize the semantic connections between them. As a result of their comparison, it was found that the semantic links between key concepts in physical and astronomical texts have fundamental differences. For each text, the average cosine proximity of some concept (for example, “atom”) with the key concepts “electron”, “proton”, “neutron”, “nucleus”, etc. can be calculated; it is one of the educational text characteristics. Conclusion. The proposed method allows to obtain objective estimates of the proximity degree of key concepts in educational texts, to study and compare semantic spaces.