1Faculty of Mathematics and Computer Science, University of Hagen, 58097Hagen, Germany, (E): philippe.tamla@fernuni-hagen.de
2Faculty of Mathematics and Computer Science, University of Hagen, 58097Hagen, Germany, (E): kouakam9210@gmail.com
3Faculty of Mathematics and Computer Science, University of Hagen, 58097Hagen, Germany, (E): bha443@proton.me
4The Energy and Resources Institute (TERI), Darbari Seth Block, India Habitat Centre, New Delhi - 110 003, India, (E): reetas@teri.res.in
5Faculty of Mathematics and Computer Science, University of Hagen, 58097Hagen, Germany, (E): matthias.hemmje@fernuni-hagen.de
This article introduces TSIAS, a framework for thematically structured information access in the exemplar domain of green ports and green shipping document collections. By analyzing a dataset of 968 journal articles (2000–24), the authors identified a significant rise in publication output from 2018, reflecting the maritime sector’s growing focus on sustainability. Using the BERT-based Topic Modeling (BERTopic) framework, 130 topics were uncovered, with the 20 most prominent—including hydrogen storage, urban planning, and shipping performance—highlighting key sustainability research priorities. A qualitative evaluation of 84 selected documents exposes limitations in current IE tools, with tools like GeneRation Of Bibliographic Data (GROBID) achieving an F1 score of 0.80, Precision of 0.86, and Recall of 0.76. These findings reveal challenges in metadata consistency and document collection organization structure, emphasizing the need for advanced Information Access (IA) solutions. The TSIAS offers modular components for metadata extraction, trend analysis, topic modelling, and visualization, providing actionable information into sustainability-focused information access. This discussed study lays the groundwork for developing advanced IA systems to aid sustainability information in green ports and green shipping.
Information extraction, Green ports, Green shipping, Topic modelling, Transformers, Named entity recognition, Taxonomy