Ontology-based semantic interoperability on the ⋆ Virtual Materials Marketplace

Abstract

The marketplace-level domain ontologies from the Virtual Materials Marketplace (VIMMP) project are presented. It is discussed

-

how a two-step ontology alignment approach can be employed for aligning the domain ontologies with the European Materials and Modelling Ontology (EMMO), the top-level ontology used by VIMMP. This is illustrated by an example from molecular modelling and simulation.

Digitalization of industrial process and product design and the uptake of industry 4.0 driven manufacturing methods is closely tied to innovations in process data technology. In this context, semantic interoperability can facilitate the integration of data with a heterogeneous provenance into a coherent framework [ 5 ].

The Virtual Materials Marketplace (VIMMP), which is under development in an ongoing Horizon 2020 project, will be an open two-sided marketplace platform, supporting the provision and acquisition of services and tools in materials modelling. For purposes of data management, ingest, and retrieval, VIMMP develops and employs a system of marketplace-level domain ontologies [ 7 ]. Thereby, knowledge representation is formalized for computational resources (domain ontology MACRO), materials modelling translation (MMTO), simulation workflows (OSMO), training (OTRAS), communication (VICO), simulation software (VISO), validation (VIVO), and model variables (VOV). An overview over this system of ontologies is given in previous work [ 7 ]; for a thorough exposition we refer to the documentation [ 8 ] and to articles that introduce the marketplacelevel domain ontologies MMTO, OSMO, and VISO specifically [ 6, 9 ].

The European Materials and Modelling Ontology (EMMO), cf. Goldbeck et al. [ 4 ], is used as a top-level ontology to facilitate interoperability between platforms. The VIMMP ontologies have been released recently [ 8 ], and for the EMMO, the full (alpha) version [ 2 ] and a simplified version (EMMO1s) used by VIMMP [ 8 ] have been made openly accessible. In Figs. 1 and 2, a knowledge graph describing a molecular model for acetylene is shown as an example. JSONLD serializations of knowledge graphs can be exchanged between VIMMP and other platforms to communicate data and metadata in a well-defined way; e.g., prospectively with the MolMod DB [ 10 ] and Bottled SAFT [ 3 ] model databases.

Acknowledgment. This work was facilitated by activities of the Innovation Centre for Process Data Technology, Kaiserslautern. Discussions with P. Gralka, C. Niethammer, G. Reina, B. Schembera, and J. Vrabec are acknowledged.

Ontology-based semantic interoperability on VIMMP 3453653)7 5

: ;< != > 9> <74 != ' > 9> 0 & ( ' ! & ( , & ' - ./ +

+ 2

+ 3453653)7 9

+ 3453653)7 + 5 " # $%% & '( )

# $* &

+ 4 ?=

References

1. CEN-CENELEC Management Centre: Materials modelling: Terminology, classification and metadata . CEN workshop agreement 17284 , Brussels ( 2018 )

2. EMMC: European Materials and Modelling Ontology ( 2020 ), https://emmc.info/emmo-info/ and https://github.com/emmo-repo/

3. Ervik , A ˚., Mej´ıa, A. , Mu¨ller, E.A.: Bottled SAFT : A web app providing SAFT-γ Mie force field parameters for thousands of molecular fluids . Journal of Chemical Information and Modeling 56 ( 9 ), 1609 - 1614 ( 2016 )

4. Goldbeck , G. , Ghedini , E. , Hashibon , A. , Schmitz , G.J. , Friis , J.: A reference language and ontology for materials modelling and interoperability . In: NAFEMS World Congress 2019 , Qu´ebec. p. NWC 19 86. NAFEMS, Knutsford, UK ( 2019 )

5. Gygli , G. , Pleiss , J.: Simulation foundry: Automated and F.A.I.R. molecular modeling . Journal of Chemical Information and Modeling 60 ( 4 ), 1922 - 1927 ( 2020 )

6. Horsch , M.T. , Chiacchiera , S. , Seaton , M.A. , Todorov , I.T. , Schembera , B. , Klein , P. , Konchakova , N.A. : Pragmatic interoperability and translation of industrial engineering problems into modelling and simulation solutions . Manuscript, Inprodat ( 2020 ). https://doi.org/10.5281/zenodo.3949803, proc . DAMDID 2020 ( accepted)

7. Horsch , M.T. , Chiacchiera , S. , Seaton , M.A. , Todorov , I.T. , Sˇindelka , K. , L´ısal, M. , Andreon , B. , Bayro Kaiser , E. , Mogni , G. , Goldbeck , G. , Kunze , R. , Summer , G. , Fiseni , A. , Bru¨ning, H., Schiffels , P. , Cavalcanti , W.L. : Ontologies for the Virtual Materials Marketplace . KI - Ku¨nstliche Intelligenz 34 ( 3 ), 423 - 428 ( 2020 )

8. Horsch , M.T. , Chiacchiera , S. , Seaton , M.A. , Todorov , I.T. , Toti , D. , Goldbeck , G.: Introduction to the VIMMP ontologies . Technical report, EMMC ASBL , Brussels ( 2020 ). https://doi.org/10.5281/zenodo.3936796

9. Horsch , M.T. , Niethammer , C. , Boccardo , G. , Carbone , P. , Chiacchiera , S. , Chiricotto , M. , Elliott , J.D. , Lobaskin , V. , Neumann , P. , Schiffels , P. , Seaton , M.A. , Todorov , I.T. , Vrabec , J. , Cavalcanti , W.L. : Semantic interoperability and characterization of data provenance in computational molecular engineering . Journal of Chemical & Engineering Data 65 ( 3 ), 1313 - 1329 ( 2020 )

10. Stephan , S. , Horsch , M.T. , Vrabec , J. , Hasse , H.: MolMod - an open access database of force fields for molecular simulations of fluids . Molecular Simulation 45 ( 10 ), 806 - 814 ( 2019 )

11. St¨obener, K. , Klein , P. , Horsch , M. , Ku¨fer, K. , Hasse , H.: Parametrization of twocenter Lennard-Jones plus point-quadrupole force field models by multicriteria optimization . Fluid Phase Equilibria 411 , 33 - 42 ( 2016 )