Handling Ambiguities in Geographic Tagging

Introduction

The vision of the Geospatial Web combines geographic data, Internet technology and social change. Geospatial applications like the IDIOM Media Watch on Climate Change facilitate geo-annotation services to refine Web pages and media articles with geographic tags.

Identifying the document’s target geographies is a rather complex task, complicated by geo/geo ambiguities (e.g. Vienna/at versus Vienna/Virginia/us) and geo/non-geo ambiguities like turkey/bird versus Turkey/country. Most approaches toward tagging the target geography therefore facilitate machine learning technologies, gazetteers, or a combination of both to identify geo-tags. The gazetteer’s size and many internal tuning parameter determine the geo-tagger’s performance and its bias towards identifying smaller geographic-entities or higher-level units. Designing a geo-tagger and choosing these parameters often involve trade-offs; improvements in one particular area does not necessarily yield better results in other areas.

The goals of this thesis are

1. designing a testcase for evaluating geo-taggers
2. implementing this testcase as a unittest
3. applying the framework to different approaches towards geo-tagging.

Table of Contents

1. Introduction
2. State of the Art
   1. Named Entity Detection
   2. Geographic Tagging
3. Method
   1. System Diagram
   2. Description of the key concepts and ideas for improving geographic tagging based on examples
4. Evaluation
   1. Description: Test Set
   2. Comparison: geoLyzard, improved Algorithm
5. Outlook & Conclusions

Literature

• E. Amitay, N. Har’El, R. Sivan, and A. Soer. “Web-a-where: geotagging web content”. In SIGIR ‘04: Proceedings of the 27th annual international ACM SIGIR conference on Research and development in information retrieval, pages 273280, New York, NY, USA, 2004. ACM.
• R. Beierheimer: “Geo Tagging of Web Resources”, Bakkalaureatsarbeit an der Technischen Universität Graz, Sept. 2006
• A. Weichselbraun: “A Utility-Testing Centered Approach for Optimizing Geo-Tagging”, draft