I am an applied scientist specialized in applications of machine learning and natural language processing for fintech and social sciences.
Previously, I was a doctoral researcher at the Jozef Stefan Institute, Artificial Intelligence Laboratory. My research areas were Machine Learning, Natural Language Processing, Semantic Web.
I defended my PhD thesis Text Annotation Using Background Knowledge in 2014.
Estimating stock price correlations using Wikipedia. At PyData London 2016 I analysed the FTSE 100 companies while at PyData Berlin 2016 I analyzed the DAX companies. [GitHub repository][Data Skeptik podcast]
Detecting novel anomalies in Twitter, a talk at PyData London 2016. [GitHub repository]
Text Stream Processing, a tutorial held together with Prof. Dr. Dunja Mladenic at the International Conference on Web Intelligence, Mining and Semantics (WIMS) 2012 [slides]
Xero recognised among top 10 most innovative companies in 2022 by Australia's Financial Review for Bank Reconciliation Predictions Retrieved 27.04.2023
Behind the tech: Using data for good to power bank reconciliation predictions Retrieved 27.04.2023
Rusu, D. et al. 2022. Transaction Data Processing Systems and Methods. International PCT application PCT/NZ2021/050151
Rusu, D. et al. 2021. Methods and systems for training attribute prediction models. Australian provisional patent application 2021903009 (International PCT application filed)
Rusu, D., Fortuna, B. and Mladenic, D. 2014. Measuring Concept Similarity in Ontologies using Weighted Concept Paths. Applied Ontology 9, no. 1, pp. 65--95.
Rusu, D., Hodson, J. and Kimball, A. 2014. Unsupervised Techniques for Extracting and Clustering Complex Events in News. Second Workshop on EVENTS: Definition, Detection, Coreference and Representation, 52nd Annual Meeting of the Association for Computational Linguistics (ACL). Baltimore, MD, USA.
Rusu, D., Fortuna, B. and Mladenic, D. 2011. Automatically Annotating Text with Linked Open Data. 4th Linked Data on the Web Workshop (LDOW), 20th World Wide Web Conference (WWW). Hyderabad, India.
RENDER - Reflecting Knowledge Diversity was a Specific Targeted Research Project (STREP) which aimed at discovering, managing and representing information on the Web through diversity-aware algorithms.
Enrycher is a service-oriented platform for enhancing text.