Entity Disambiguation

This lecture covers the concept of Entity Disambiguation, which involves linking text mentions to entries in a knowledge base. It discusses challenges like homonyms and synonyms, and methods such as Label Propagation and Link Prediction. The instructor explains the importance of coherence in entity graphs and the use of Personalized PageRank for disambiguation. The lecture also explores the approach of finding the most likely valid concept candidate linked to a mention, considering factors like total support and popularity measures. Various results from different disambiguation methods are presented, along with a quiz on entity disambiguation and named entity recognition.