INEX 2010 Link-The-Wiki Track

Overview

The Link-the-Wiki track aims to produce a standard procedure and metrics for the evaluation of link discovery between documents. Given an unlinked document collection, the task is to analyse the text and recommend a set of incoming and outgoing links from anchor-text to the Best Entry Point (BEP) in other documents in the collection. This means that anchor text will be linked to a specific position in the target document - the best entry point for starting to read the referenced material from.

Going beyond traditional text document analysis, both file level and element level link discovery will be evaluated.

For link discovery, participants will be given a set of existing Te Ara Encyclopedia documents. The task is to discover incoming and outgoing links. This task requires collaborative assessment by participants in the INEX tradition. An assessment GUI will be used that will allow for efficient assessment of links.

The task is similar to last year's Link-Te-Ara task, however the collection will be a fresh dump of the Te Ara Encyclopedia

2010 Data

Go to the Te Ara Homepage
Te Ara 2010 Document Collection (the INEX task is to link this to itself)
Te Ara Taxonomy leading to Themes.

2009 Data

2009 anchor-to-BRP topics (33 participant submitted)
2009 topics (5000 randomly selected)

2008 Data

2008 topics (document-to-document)
2008 topics (anchor-to-BEP)

Software

Link-the-Wiki run Validation Tool
Link-the-Wiki run Validation Tool Resources (sample runs, instructions, DTD, etc.)

Schedule

22/Oct/2010Submission deadline for Link-the-Wiki runs
5/Nov/2010Submission deadline for Link-the-Wiki assessments
12/Nov/2010Release of Link-the-Wiki evaluation results

Organisers

Shlomo Geva
Queensland University of Technology
s.geva@qut.edu.au

Andrew Trotman
University of Otago
andrew@cs.otago.ac.nz