Change search
ReferencesLink to record
Permanent link

Direct link
On a New Model and Algorithm to Estimate Geographical Similarity between Document and Query in GIR System
KTH, School of Architecture and the Built Environment (ABE), Urban Planning and Environment, Geoinformatics (closed 20110301). (Geoinformatics)
2008 (English)In: The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2008Conference paper (Other academic)
Abstract [en]

The usage of single geographical footprint model in existing Geographical Information Retrieval (GIR) system will cause manyproblems like overestimation or underestimation of the geographical scopes for documents to search. To be honest, the singlegeographical footprint model is not applicable in modern GIR system although it is simple, fast and widely applied today. In order toimprove the quality of answers given to a spatial query, a new model as well as a dedicated algorithm will be proposed to study thegeographical information attached to documents. Besides, a dedicated algorithm based on network graph, which is inspired by theGoogle PageRank and Bayesian network theory, is also invented to estimate the geographical similarity between document€d andquery€q based on this new model. We believe that the new model and algorithm proposed here could better estimate thegeographical similarity between document and query for a GIR system. Because it not only consider the geometric adjacency ofdocument and query, but also take into account the importance of each places towards the hosting documents. By using the newmodel, the geographical scope of document could be better studied, and thus improve the quality of answers from GIR systemconcerning the spatial query.

Place, publisher, year, edition, pages
Keyword [en]
Spatial Information Science, GIS, Information, Service, Information Retrieval, Algorithm, Retrieval Model
National Category
Computer and Information Science
URN: urn:nbn:se:kth:diva-89121OAI: diva2:502711
The XXI Congress of ISPRS. Beijing, China. 3-11 July 2008
QC 20120509Available from: 2012-02-14 Created: 2012-02-14 Last updated: 2012-05-09Bibliographically approved

Open Access in DiVA

No full text

Search in DiVA

By author/editor
Lin, Xing
By organisation
Geoinformatics (closed 20110301)
Computer and Information Science

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 16 hits
ReferencesLink to record
Permanent link

Direct link