Discovering the Hidden Cross-Dataset Links in Data.gov
暂无分享,去创建一个
Data mash-ups unleash users from the limit of accessing one
dataset at a time, and enable a broader view based on the
integrated data. A key challenge to mashing up Data.gov datasets
lies in the fact that cross-dataset links are rarely published
explicitly by dataset owners, making it hard for users to find
related datasets for building mash-ups. In this paper, we show
several types of hidden cross-dataset links found in Data.gov and
explain how they can be obtained using semantic technologies in
Rensselaer's Linking Open Government Data project.
[1] Xiaolong Li,et al. An Overview of Microsoft Web N-gram Corpus and Applications , 2010, NAACL.
[2] Lise Getoor,et al. Collective entity resolution in relational data , 2007, TKDD.
[3] James A. Hendler,et al. TWC data-gov corpus: incrementally generating linked government data from data.gov , 2010, WWW '10.