Toward an enhanced Arabic text classification using cosine similarity and Latent Semantic Indexing

The Faculty of Systems and Information Engineering of the National University of San Martin, is constantly developing projects looking for new forms of innovation in order to meet needs and improve the quality of life of the university community and society in general. The world of research is very broad, researching a certain subject that was previously worked on is common, therefore it is difficult to determine the similarity in a manual way, because there is no technological tool that automates and systematizes these processes, This generates uncertainty in the world of research, researchers are unaware of the content of previous research, which can therefore end up generating duplication of the same. The following research project entitled "Computer system based on the model of vector space for the identification of the degree of similarity of thesis projects in the School of Systems and Information Engineering of the UNSM-T", developed with funding from the National University of San Martín in 2018, having as objective to implement a technological tool that allows to identify the degree of similarity of thesis projects in the School of Systems and Information Engineering of the UNSM-T, with this objective, the research question is the following, how to identify the degree of similarity of the thesis projects in the faculty of systems engineering and computer science of the UNSM-T? ", the answer to this question is done through a study of applied type and level of pre-experimental research. In addition, a pre and posttest was designed, which allowed knowing the situation of a before and after the implementation of the computer system (http://jauregui.infinitumperu.com/). The answers that were obtained regarding the quality of the search results (theses recovered through a query) indicate that 59% of the respondents rated it as "Good" and 41% as "Very good". Taking into account the aforementioned, it is recommended to massif the use of the proposal for the identification of the degree of similarity of thesis projects in the FISI, in such a way as to avoid duplication or develop a thesis project with a high degree of similarity with an already developed and sustained in that faculty.