Description of the BOUN System for the Trilingual Entity Detection and Linking Tasks at TAC KBP 2017

This paper describes the BOUN system’s participation in the Trilingual Entity Detection and Linking (EDL) track in 2017 TAC Knowledge Base Population (KBP) challenge. EDL is important aspect of text analysis and various tasks like sentiment analysis and opinion mining can benefit from discovering which named entities are mentioned in the context. For this challenge, we built a simple candidate generator and applied a Maximum Entropy-based approach for named entity linking. Our system has achieved an F1 score of 52.5% in the “strong typed all match” metric on the 2017 evaluation set.