To make machines able to find people, organisations, events and other named-entities in text we need to define ways to separate which tokens in the text that are part of a named-entity and which are not. In this presentation I present a solution to the problem of Named-Entity Chunking for Norwegian text using Support Vector Machines. As a basis for Named-Entity Recognition, this can enable us to do advanced analysis of the actors in a text and their relationship with it and each other.

Watch the webcast replay >>