Improvement Recall of Person Name Disambiguation on the Web People Search by TwoStage Clustering

@inproceedings{Ikeda2009ImprovementRO,
  title={Improvement Recall of Person Name Disambiguation on the Web People Search by TwoStage Clustering},
  author={Masaki Ikeda and Shingo Ono and Issei Sato and Minoru Yoshida and Hiroshi Nakagawa},
  year={2009}
}
This research proposes the application of semi-supervised learning to unsupervsed clustering. There are two criteria of cluster evaluation, or precision and recall. Precision is the ratio of true datas in the result cluster and recall is the ratio of true datas the result cluster has to all true data. In previous work, the selection of feature types enables to make high precision clusters, but these features are too sparse to imporve recall. On the otherhand, there are features that has poor… CONTINUE READING