Article accepted to ACM Computing Surveys

We are happy to announce that our paper “Combining Machine Learning and Semantic Web: A Systematic Mapping Study” has been accepted to ACM Computing Surveys in February 2023 [1]. We were pleased to experience the great interest of the community in this work: from its publication less than 2 months ago, on the 9th of March, the paper was downloaded more than 600 times!

This article is motivated by the rise of systems using semantic resources and machine learning approaches solving a variety of tasks. We surveyed 476 papers that we coined as Semantic Web Machine Learning Systems (SWeMLS). The main results consist of 1) a trends landscape in terms of maturity, characteristics of the different components, and 2) a classification system for SWeML Systems. The pre-print can be accessed here. 

Among the results are also an overview of processing patterns (cf. Fig 1.) derived from the architectural characteristics of SWeML Systems. These patterns are inspired by the work of Van Harmelen and Ten Teije [2] and reuse their existing 11 patterns, but do not focus on the task but on the architectural flow. They describe processing components, in- and outputs (either data or symbolic resources) of a given SWeML System. Additionally, we identified 33 new patterns from our surveyed papers.

The work was an international collaboration between Semantic Web Company, TU Wien, WU Wien, University of Vienna, University of Mannheim, and VU Amsterdam and also part of the FFG-funded OBARIS project.

[1] A. Breit, L. Waltersdorfer, F. J. Ekaputra, M. Sabou, A. Ekelhart, A. Iana, H. Paulheim, J. Portisch, A. Revenko, A. ten Teije, F. van Harmelen “Combining Machine Learning and Semantic Web: A Systematic Mapping Study,” ACM Computing Surveys, 2023

[2] Van Harmelen, F., & Teije, A. T. (2019). A boxology of design patterns for hybrid learning and reasoning systems. arXiv preprint arXiv:1905.12389.