Session Number

Parallel Session 3C

Description

As the amount of content in the engineering space is constantly growing it is becoming more and more difficult to retrieve relevant information. In this context the IET has commissioned a project to evaluate the feasibility of providing a semantic enrichment service based on the terms contained in the Inspec Thesaurus. The possibility of automated tagging of a variety of content types including traditional academic publishing data (e.g. journal abstracts), commercial product information (e.g. product catalogues) and social media (e.g. twitter feeds) has been explored as well as the auto generation of new terms for the Inspec thesaurus. We have used recognized semantic data modelling techniques to create a working ontology for a subset of the Inspec thesaurus and then used this ontology to entity extract and tag the names of people, organisations and their locations, along with Inspec thesaurus terms themselves, on all items of content run through the system. In addition we have built a number of applications to test the validity and accuracy of this approach as well as to explore how such a system would cope with the volume of future indexing requirements. We are continually evaluating our approach via on-going market research and user testing.

Share

COinS
 
Jun 3rd, 12:00 AM

Beyond Inspec - Building a Semantic Enrichment Service to Explore How the IET can Better Support Academic and Commercial Researchers

As the amount of content in the engineering space is constantly growing it is becoming more and more difficult to retrieve relevant information. In this context the IET has commissioned a project to evaluate the feasibility of providing a semantic enrichment service based on the terms contained in the Inspec Thesaurus. The possibility of automated tagging of a variety of content types including traditional academic publishing data (e.g. journal abstracts), commercial product information (e.g. product catalogues) and social media (e.g. twitter feeds) has been explored as well as the auto generation of new terms for the Inspec thesaurus. We have used recognized semantic data modelling techniques to create a working ontology for a subset of the Inspec thesaurus and then used this ontology to entity extract and tag the names of people, organisations and their locations, along with Inspec thesaurus terms themselves, on all items of content run through the system. In addition we have built a number of applications to test the validity and accuracy of this approach as well as to explore how such a system would cope with the volume of future indexing requirements. We are continually evaluating our approach via on-going market research and user testing.