Counting Information
from Text


CINEX is the first full-fledged system for extracting counting information from text. Notable examples of extracted counts include "Humboldt never married" and "Katia and Thomas Mann had six children". Such counting knowledge enables a wide range of applications, such as aggregate querying and knowledge base quality estimator. CINEX is presented at the International Semantic Web Conference (ISWC) 2018 in California, USA, and was developed by Paramita Mirza, Simon Razniewski, Fariz Darari, and Gerhard Weikum. CINEX has been empirically evaluated over real datasets, that is, Wikipedia and Wikidata, and is able to assert the existence of 2.5 million counting facts for 110 relations.

Ask CINEX



2018 | Max Planck Institute for Informatics and Universitas Indonesia