Entity Linking

Identifying named entities in queries and linking them to the corresponding entry in the knowledge base is known as the task of entity linking in queries (ELQ). Given a query q, return one or multiple interpretations of the query, each interpretation consists of a set of mention-entity pairs.

Entity retrieval is a core building block of semantic search. Given a search query, entity retrieval is the task of returning a ranked list of entities from an underlying knowledge base.

The following ELQ methods are implemented in Nordlys:

  • CMNS: Thee baseline method that performs entity linking based on the overall popularity of entities as link targets, i.e., the commonness feature [Hasibi et al., 2015].
  • LTR-greedy: The recommended method (with respect to both efficiency and effectiveness) by Hasibi et al. [Hasibi et al., 2016], which employs a learning-to-rank model with various textual and semantic similarity features. Please note that the implemented method in Nordlys is slightly different from the one presented in [Hasibi et al., 2016] (i.e. the features and index). The corresponding files for this method are under data/el. Specifically:
    • config_ltr.json holds the config file
    • model.txt is the trained model

References

  • Faegheh Hasibi, Krisztian Balog, Svein Erik Bratsberg. 2015. Entity Linking in Queries: Tasks and Evaluation. In: ACM SIGIR International Conference on the Theory of Information Retrieval (ICTIR ’15). [PDF]
  • Faegheh Hasibi, Krisztian Balog, Svein Erik Bratsberg. 2017. Entity Linking in Queries: Efficiency vs. Effectiveness. In: 39th European Conference on Information Retrieval (ECIR ’17). [PDF]
  • Paul Ogilvie and Jamie Callan. 2003. Combining Document Representations for Known-Item Search. In: 26th annual international ACM SIGIR conference on Research and development in information retrieval (SIGIR ‘03).