att_abstract={{Knowledge bases such as DBpedia are often used in an exploratory mode, in order to "tell you more" information related to a given fact. Since much information related to a DBpedia fact (corresponding to a Wikipedia Infobox attribute name-value pair) is present in the textual part of the Wikipedia page, we focus on the problem of role linkage, which aims at finding and linking (possibly different) entities in the Infobox and text parts across time that have the same semantics.  Role linkage has a spatial dimension of linking occurrences of a given entity that have the same semantics within a Wikipedia page version, and a temporal dimension of linking occurrences of possibly different entities that have the same semantics across page versions. We formalize the goal of role linkage in Wikipedia as producing a ladder, where the rungs capture the spatial dimension, and the legs capture the temporal dimension. We design and implement algorithms to efficiently extract ladders from the revision history of a Wikipedia page, making use of supervised learning techniques to account for the different kinds of edits that happen in Wikipedia pages. We empirically demonstrate the strengths of the ladder method using four Wikipedia based datasets and ground truth labels obtained from AMT.}},
	att_categories={C_BB.1, C_NSS.2, C_IIS.5},
	att_copyright_notice={{(c) ACM, 2017. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in 20th International Workshop on the Web and Databases (WebDB 2017) {{, 2017-05-14}}.
	author={Divesh Srivastava and Siarhei Bykau and Jihwan Lee and Yannis Velegrakis},
	institution={{20th International Workshop on the Web and Databases (WebDB 2017)}},
	title={{"Tell me more" using Ladders in Wikipedia}},