In the scientific digital libraries, some papers from different research\ncommunities can be described by community-dependent keywords even if they share\na semantically similar topic. Articles that are not tagged with enough keyword\nvariations are poorly indexed in any information retrieval system which limits\npotentially fruitful exchanges between scientific disciplines. In this paper,\nwe introduce a novel experimentally designed pipeline for multi-label\nsemantic-based tagging developed for open-access metadata digital libraries.\nThe approach starts by learning from a standard scientific categorization and a\nsample of topic tagged articles to find semantically relevant articles and\nenrich its metadata accordingly. Our proposed pipeline aims to enable\nresearchers reaching articles from various disciplines that tend to use\ndifferent terminologies. It allows retrieving semantically relevant articles\ngiven a limited known variation of search terms. In addition to achieving an\naccuracy that is higher than an expanded query based method using a topic\nsynonym set extracted from a semantic network, our experiments also show a\nhigher computational scalability versus other comparable techniques. We created\na new benchmark extracted from the open-access metadata of a scientific digital\nlibrary and published it along with the experiment code to allow further\nresearch in the topic.\n