Regresar

Case of study of latent Dirichlet assignment method (LDA) in literary review processes

Abstract:

The literary review processes have used multiple methods for the search and classification of studies on various topics. This paper presents an application of the Latent Dirichlet Assignment Method (LDA) to support the process of classification and filtering of studies. Searches were carried out in the WOS, Scopus, Google Scholar, and IEEE databases in housing recommendation systems. The literary review process was carried out considering the following sequence: i) Search in databases by keywords; ii) clean text exploration; iii) construction of the model using LDA, and iv) analysis of article classification. Five topics were obtained and grouped in organized tables with their respective titles and citations for each article. The set of steps applied allows establishing some key ideas to consider in designing the systems of recommendations for housing.