Link-the-Wiki: At INEX we are using the Wikipedia collection - about 5 GBytes consisting of 660,000 documents, in XML format. The document set is extensively hyperlinked. The Link-the-Wiki task aims at evaluating the state of the art in automated discovery of document hyperlinks. The objective of the task is to provide an evaluation forum and a set of standard tasks and corresponding achievable results. We aim to create a reusable resource for evaluating and comparing different state of the art systems and approaches to automated link discovery. More specifically, given a new orphan wikipedia document, the task is to analyse the text and recommend a set of incoming and outgoing links from/to anchor text in the existing collection. Going beyond traditional text document analysis, in the context of INEX we aim to operate at the XML element level. This means that anchor text or anchor elements will link not only to a related document, but to a specific XML element within, or to the best entry point for starting to read the referenced material from.

Schedule