Huge congratulations to LT4SG PhD student Menasha Thilakaratne for submitting her thesis on her amazing research on ‘Literature Based Discovery (LBD)’. Menasha joined LT4SG in 2017 and established LBD research area within LT4SG. She has published her research in many highly-ranked venues.
In Search of a Common Thread: Enhancing the LBD Workflow with a view to Widespread Applicability
Literature-Based Discovery (LBD) research focuses on discovering implicit new knowledge linkages from existing scientific facts to provide impetus to research progress and increase research productivity. Despite the significant progress of LBD research, there are several open issues and shortcomings in the previous studies that are in critical need to be addressed. The overarching goal of this thesis is to fill these gaps with high precision, to not only enhance the discovery component as most of the previous LBD studies, but also to shed light on several new directions to further strengthen the existing understanding of the LBD workflow. In accordance with this goal, the thesis aims to enhance the LBD workflow within sight of ensuring its widespread applicability.
The aforementioned intention of widespread applicability is twofold. Firstly, it signifies the flexibility of the proposed solutions to be utilised in a diverse range of other problem settings. These problem settings do not necessarily imply application areas closely related to the LBD context (e.g., drug development, collaboration recommendation), but also a wide range of other problems (e.g., product innovation, personalisation) that deviate from the scope of LBD. The main motivation for this is that the intrinsic objective of LBD research, which is discovering novel linkages from text corpora, could be broadly applicable in a vast range of problem settings. Thus, extricating domain-specialised inferences and developing easily adaptable LBD models assist in formulating expeditious solutions in a diverse range of problem settings. While enlarging the applicability of the LBD models, this also increases their dependability.
- Thilakaratne, M., Falkner, K., Atapattu, T. Garbage in-garbage out? an empirical look at information richness of LBD input types . ACM/IEEE Joint conference on Digital Libraries, Aug 2020
- Thilakaratne, M., Falkner, K., Atapattu, T. Information extraction in digital libraries: first step towards portability of LBD workflow . ACM/IEEE Joint conference on Digital Libraries, Aug 2020
- Thilakaratne, M., Falkner, K., Atapattu, T. Connecting the dots: Hypothesis generation by leveraging semantic shifts, PAKDD 2020, Singapore.
- Thilakaratne, M., Falkner, K., Atapattu, T. A systematic review on literature-based discovery: General overview, methodology & statistical analysis , ACM Computing Surveys (CSUR), 129, Dec 2019
- Thilakaratne, M., Falkner, K., Atapattu, T. A systematic review on literature-based discovery workflow , PeerJ Computer Science 5:e235, Nov 2019.
- Thilakaratne, M., Falkner, K., Atapattu, T. Automatic Detection of Cross-Disciplinary Knowledge Associations. Student Research Workshop of 56th Annual Meeting of the Association for Computational Linguistics, Melbourne, Australia, July 2018