To manage and measure the performance of scientific research at the university, managers or policymakers need synthetic indicators that are cleverly grouped in several indicators which aim to offer to the leaders the necessary tools, so as to improve scientific research. The governance of information system has posed a serious challenge for the leadership of universities especially through the use of decision aid. In order to improve the information system, especially scientific research, a study of automatizing curriculum vitae for researchers of different disciplines belonging to various research laboratories is discussed in this paper. The use of natural language processing with data mining classifier Decision Tree is presented in order to predict the field of work of each researcher. The choice of Decision Tree classifier among One Rule classifier and Naive Bayes classifier is not arbitrary; it is chosen by comparing performance metrics- such as Precision, Recall, F-measure, Correctly Classify Instance, Incorrectly Classify Instance, Kappa Statistic, Root Mean Squad Error, Relative Absolute Error, and Root Relative Squad Error. [ABSTRACT FROM AUTHOR]