The article is devoted to testing the method of automatic calculation of the emotional characteristics of contexts on textual examples from «Tuḥfat al-nuẓẓār fī ġarā’ib al-amṣār wa ‘ajā’ib al-asfār» by Ibn Battuta. The study is one of the first attempts to examine the emotional coloring of medieval Arabic texts using this method, originally developed for the study of modern media content. To determine the tone of the text, the CAMeL Tools algorithm was applied to analise the contexts from the document containing references to Christians. Manual verification of the results obtained using the algorithm demonstrated a fairly high degree of efficiency of this algorithm for the analysis of the medieval Arabic text, since the tone was correctly determined in 72 % of the contexts. The study showed that the emotional coloring of the contexts about Christians (naṣа̄rа̄) in Ibn Battuta’s text is neutral in most cases. Further training of the neural network on the corpus of medieval texts and improving the efficiency of the algorithm will make it applicable to large amounts of data for the purpose of determining the emotional coloring of lexemes numbering hundreds and thousands of word usages.