ABSTRACT. In this tutorial we discuss how Conditional Random Fields can be applied to knowledge base population tasks. We are in particular interested in the cold-start setting which assumes as given an ontology that models classes and properties relevant for the domain of interest, and an empty knowledge base that needs to be populated from unstructured text. More specically, cold-start knowledge base population consists in predicting semantic structures from an input document that instantiate classes and properties as dened in the ontology. Considering knowledge base population as structure prediction, we frame the task as a statistical inference problem which aims at predicting the most likely assignment to a set of ontologically grounded output variables given an input document. In order to model the conditional distribution of these output variables given the input variables derived from the text, we follow the approach adopted in Conditional Random Fields. We decompose the cold-start knowledge base population task into the specic problems of entity recognition, entity linking and slot-lling, and show how they can be modeled using Conditional Random Fields.

ABSTRACT. This is a quick survey about efficient search on a text corpus combined with a knowledge base. We provide a high-level description of two systems for searching such data efficiently. The first and older system, Broccoli, provides a very convenient UI that can be used without expert knowledge of the underlying data. The price is a limited query language. The second and newer system, QLever, provides an efficient query engine for SPARQL+Text, an extension of SPARQL to text search. As an outlook, we discuss the question of how to provide a system with the power of QLever and the convenience of Broccoli.