In the domain of historical data interoperability [1] one of the major points under discussion is about the possibility of sharing a common ontology for modelling the whole of human activities in the past. The ontology should also describe the method used to extract from the sources the information expressed in the data. To cope with this issue the ​symogih.org project (Système modulaire de gestion de l’information historique) —started in 2008 with the aim of producing a virtual research environment (VRE) for collaborative data production— developed a shared vocabulary about specific sub-domains of historical research providing data interoperability among research projects hosted in the platform. The different parts of the VRE took also into account the different epistemological levels of historical information production, documenting the origin in the sources [2]. This perspective has been broadened since 2013 with the aim of sharing data produced in the ​symogih.org virtual research environment as Linked open data and integrating them with other resources available on the semantic web using RDF technologies [3]. For this purpose the project adopted the CIDOC CRM as a conceptual framework which is not only a standardized (ISO 21127:2014) “formal ontology intended to facilitate the integration, mediation and interchange of heterogeneous cultural heritage information” but also a generic data model designed with holding the aim to “maintain and support a global knowledge network” [4]. Due to its high level of abstraction, the CRM can be applied to different research areas by creating specific extensions. To support this process, the LARHRA has developed an Ontology management environment (OntoME) as an online application that offers research projects the ability to manage their ontologies (data models) in a collaborative and open way [5]. OntoME is a key element in the international ​Data for History (DfH) consortium (dataforhistory.org), founded in 2017 at the initiative of the LARHRA with the aim of improving geo-historical data interoperability in the semantic web [6]. In this paper I will present the workflow by which a project can express its own data model using OntoME, then align the model with the CIDOC CRM and with specific extensions designed to model sub-domains of historical research and finally use an appropriate application profile to publish own data on the semantic web. In particular, I will address the question of the different epistemological levels of data produced by historical research and present a model for documenting information extraction from sources and the process of integration of existing data. [1] Meroño-Peñuela Albert, Ashkpour Ashkan, van Erp Marieke, Mandemakers Kees, Breure Leen, Scharnhorst Andrea, Schlobach Stefan, van Harmelen Frank, « Semantic Technologies for Historical Research: A Survey », in Semantic Web – Interoperability, Usability, Applicability (IOS Press) 6(2015): 539-564. [2]​ ​http://symogih.org/?q=documentation [3]​ ​http://symogih.org/?q=rdf-publication​ – Cf. Beretta Francesco. L'interopérabilité des données historiques et la question du modèle : l'ontologie du projet SyMoGIH. ​Enjeux numériques pour les médiations scientifiques et culturelles du passé, Paris, Presses Universitaires de Paris Nanterre, 2017, 87-127. [4] Martin Doerr et al., ‘The Dream of a Global Knowledge Network: A New Approach’, Journal on Computing and Cultural Heritage, vol. 1, no. 1 (2008). [5] ​http://ontome.dataforhistory.org/ [6] http://dataforhistory.org/meetings .