Quantifying the accuracy of relational statements in Wikipedia: a methodology

The Perseus Project at Tufts University produces tools to enhance the study of humanities texts. Perseus’ new named-entity browser lets users browse an index of references to people, places, and dates within a work. A more advanced version of this same tool would enable users to navigate a work by specifying relations between two entities. Such tools require sets of relational data to function. This paper discusses a methodology for evaluating data sets obtained from Wikipedia and its potential. Categories and Subject Descriptors H.3.3 [Information Storage and Retrieval]: Information Search and Retrieval; H.5.4 [Information Systems Applications]: Hypertext/HypermediaArchitectures General Terms design, experimentation Keywords Wikipedia, named-entity recognition, link analysis

• Type: Conference
• Year: 2006
• Where: JCDL
• Authors: Gabriel Weaver, Barbara Strickland, Gregory Crane
• Keywords: Browser Lets Users, JCDL 2006, Tools Require Sets, University Produces Tools