LINQS

STATISTICAL RELATIONAL LEARNING GROUP @ UMD



 

Collective Entity Resolution in Familial Networks

Pigi Kouki, Jay Pujara, Christopher Marcum, Laura Koehly, Lise Getoor
IEEE International Conference on Data Mining (ICDM) - 2017
Note: To Appear  
Download the publication : kouki-icdm17.pdf [669Ko]  
Entity resolution in settings with rich relational structure often introduces complex dependencies between coreferences. Exploiting these dependencies is challenging – it requires seamlessly combining statistical, relational, and logical dependencies. One task of particular interest is entity resolution in familial networks, which is crucial for applications such as tracking disease contagion, understanding genetic inheritance, and performing census surveys. In this problem setting, we are provided partial, and possibly unreliable, views of a familial network from the perspective of multiple, different family members. The modeling goal is to match entities across these incomplete and noisy views. Here we design a model that incorporates statistical signals, such as name similarity, relational information, such as sibling overlap, and logical constraints, such as transitivity and bijective matching, in a collective model. We show how to integrate these features using probabilistic soft logic, a scalable probabilistic programming framework. In experiments on real-world data, our model significantly outperforms state-ofthe- art classifiers that use relational features but are incapable of collective reasoning.

BibTex references

@InProceedings{kouki:icdm17,
  author       = "Kouki, Pigi and Pujara, Jay and Marcum, Christopher and Koehly, Laura and Getoor, Lise",
  title        = "Collective Entity Resolution in Familial Networks",
  booktitle    = "IEEE International Conference on Data Mining (ICDM)",
  year         = "2017",
  note         = "To Appear",
}

Other publications in the database