In the framework of RADNEXT Work Package 3 (WP3), the Transnational Access (TA) portal was created for the management of the submission process of experimental proposals. One of the WP3 research activities, and the topic of an ongoing PhD project, is to support the assessment of TA requests. Using Natural Language Processing (NLP) techniques, we aim to provide automatic assistance for all the interested stakeholders. That includes support for both the users during the submission process and reviewers and User Selection Panel (USP) members during the project-selection period. This poster introduces the machine learning and data-driven tools used for these goals. Some initial experiments, accomplished tasks and created software are presented in order to notify the RADNEXT network about the current status and advances of this research and provide a baseline for subsequent discussion related to such activities. We take advantage of the Open Peer Review (OPR) movement to gather High-Energy Physics (HEP)-related data to build and train custom Machine Learning (ML) models able to provide initial evaluation of experimental proposals. The presented research is highly innovative, since NLP-based processing is mostly used in the field of computer science and human sciences - and not necessarily High-Energy Physics. Future plans for the use of ML methods are presented - both in the framework of RADNEXT-related activities and resulting PhD thesis research.