Abstract: The SDG Hub is a platform that seeks to provide access to researchers on relevant articles that speak to the sustainable development goals (SDGs). It is critical to have a robust and user-friendly document search engine that allows researchers to effectively search for relevant SDG information. To achieve this task, a meta-data dataset was provided by the Albert Luthuli institution at the University of Pretoria. We propose a study on the document search of the articles using machine learning techniques. Firstly, we apply an SVD model to retrieve topics based on the word similarity measure. Secondly, we apply an LDA model to study the homogeneity of the documents and the correlation of topics to article themes. The results were evaluated and presented using visualization tools.