Wikipedia has become one of the best sources for creating and sharing a massive volume of human knowledge. Much effort has been devoted to generating and enriching the structured data by automatic information extraction from unstructured text in Wikipedia. Most, if not all, of the existing work share the same paradigm, that is, starting with information extraction over the unstructured text data, followed by supervised machine learning. Although remarkable progresses have been made, this paradigm has its own limitations in terms of effectiveness, scalability as well as the high labeling cost. We present WiiCluster, a scalable platform for automatically generating infobox for articles in Wikipedia. The heart of our system is an effective cluster-then-label algorithm over a rich set of semi-structured data in Wikipedia articles: linked entities. It is totally unsupervised and thus does not require any human label. It is effective in generating semantically meaningful summarization for Wikipedia articles. We further propose a cluster-reuse algorithm to scale up our system. Overall, our WiiCluster is able to generate nearly 10 million new facts. We also develop a web-based platform to demonstrate WiiCluster, which enables the users to access and browse the generated knowledge.
[1]
Mitsuru Ishizuka,et al.
Exploiting Syntactic and Semantic Information for Relation Extraction from Wikipedia
,
2006
.
[2]
Mitsuru Ishizuka,et al.
Subtree Mining for Relation Extraction from Wikipedia
,
2007,
NAACL.
[3]
Greg Hamerly,et al.
Learning the k in k-means
,
2003,
NIPS.
[4]
Chris H. Q. Ding,et al.
Infobox suggestion for Wikipedia entities
,
2012,
CIKM '12.
[5]
Hanghang Tong,et al.
The Links Have It: Infobox Generation by Summarization over Linked Entities
,
2014,
ArXiv.