An investigation is presented through the processing of natural language and text analysis found in the abstracts of the educational publication Profundidad of the Faculty of Administrative and Economic Sciences at the Universidad Francisco de Paula Santander Ocaña. The indicators analysed are magnitude, year of publication, importance weighting and sentiment index, so that it can be determined if the sentiment index determines the percentage of citations of the articles in importance with reference to the other indicators. The number of articles in the publication is 107 articles and a sample of 13 abstracts selected among those with the highest index and similar in score. The analysis is performed using the words in Spanish, the official language of the publication, since the translation may vary the score due to its meaning.