/
Business Data
Work Phone: 
+359878992829
Personal Bio

Dr. Georgi Georgiev has specialised in advanced text analytics with a focus on machine learning based models and overall software architectures and solution methodologies for enterprise level semantic annotation and search solutions. Georgiev leads the text analysis product development and professional services at Ontotext and manages semantic publishing projects for organisations such as the BBC and PressAssociation. He has a solid leading role in the Semantic Annotation and Search division of Ontotext and a strong commitment to combine scientific and research excellence with productive industrial and government oriented solutions. His interests include machine learning product development and adaptation, team leadership, enterprise software architectures, and contemporary management techniques.

Presentation title: 
Semantic Solutions for Scientific and Academic Publishers
Presentation description: 
Our talk will discuss how to bring value and benefit by applying Semantic technology to the domain of Scientific and Academic Publishing. Building on Ontotext’s rich experience with leading media enterprises like BBC and Euromoney, we delivered a semantic solution for The Institution of Engineering and Technology (IET). The role of semantics in the IET's publishing process is to automate and scale-up parts of their current publishing workflow so that they can deliver more relevant high-value content across customers and channels. Their archive aggregates nearly 1,000 publishers and reaches more than 60 million documents, collected since 1969. The solutions we present is based on Ontotext’s Dynamic Semantic Publishing Platform, that is built on SOA principles for LOD management and provides semantic enrichment and personalisation at scale. We will outline the challenges we had faced in implementing and integrating semantic publishing solution in IET’s complex scenario in various stages as: • building coherent semantic publishing technical and information architectures; • integrating main software components that generate and consume semantic metadata; • applying metadata enrichment tools that consume “Inspec” - the world-leading thesaurus in the engineering domain, built in hundreds of scientific craft days. Finally we will then dive into quality, scalability, reliability and availability challenges at IET as they appeared to be common for many semantic technology applications in enterprise environment. IET initial target for high-quality automatic enrichment starts from 850,000 pieces of content per year, only 5% of which are planned to be manually verified by subject matter experts.