Datoptron is a spinoff company of the National Technical University of Athens in the area of Artificial Intelligence. We leverage 20 years experience in Semantic Technologies, Machine Learning, and Crowdsourcing.
Learn More
Analysing information requires structured and accessible data for best results. Data transformation procedures enable organisations to adapt the representation and format of raw data and structure them depending on their needs. Datoptron sets in place well-defined and customisable procedures for mapping between different schemata.
Datoptron offers situation-appopriate solutions for making data uniform and interoperable. We use Natural Language Process, named entity extraction and pattern matching techniques to automatically link and enrich your data with trusted external web resources. We validate the results using a human-in-the-loop approach.
How can data be published as Linked Data? How can you unearth the knowledge underlying the data, identify hidden patterns, and discover what you are looking for? Datoptron's expertise in both relational and graph databases ensures that the appropriate approach for storing and quering your data will be followed.
Our platform development expertise makes sure that the capabilities enabled by state-of-the-art data technologies are served to the end user in a functional and appealing way. We build on well-defined Application Programming Interfaces and deliver user-griendly interfaces using UX design best practices.
We combine the power of state-of-the-art Artificial Intelligence tools with experts’ and crowd human intelligence for achieving high quality results and for maximising engagement. We organise crowdsourcing campaigns and exploit human feedback to gain helful insights, fine-tune and improve the automatic processes as well as to constructfully involve various audiences.
We build situation-appropriate standards in all the above-mentioned steps, making use of well-founded schemata, controlled vocabularies, and ontologies to represent and manage data and metadata. We opt for open standards, wherever possible, and advocate the FAIR (Findable, Accessible, Interoperable and Reusable) principles.
Leveraging the experience of 20 years working in the fields of Semantic Technologies and Machine Learning and the participation in more than 30 R&D European-funded projects, Datoptron has developed a number of tools.
View More
Metadata Interoperability Platform (MINT) was designed and developed to facilitate aggregation initiatives for cultural heritage content and metadata in Europe. More than 15 million cultural records have been aggregated to Europeana through MINT
SAGE (Semantic Annotation & Generation of Enrichments) is a web-based tool for generating, enriching, validating, searching and publishing data as a knowledge graph.
Is a mapping language, developed inhouse, and is used to traasnform any kind of data to RDF. D2RML has been used in many European R&D projects that work with Linked Data technologies.
CrowdHeritage is an open platform where cultural heritage institutions can share their collections’ metadata that need a fix or enrichment, and everybody can contribute to improve them.
We are a group of talented people including researchers, software developers, ontology engineers and machine learning experts with long experience working with standards.
Let's Work Together