A descriptive model for predicting popular areas in a web map

  • Authors:
  • Ricardo García;Juan Pablo De Castro;María Jesús Verdú;Elena Verdú;Luisa María Regueras;Pablo López

  • Affiliations:
  • Department of Signal Theory, Communications and Telematics Engineering, Higher Technical School of Telecommunications Engineering, University of Valladolid, Valladolid, Spain;Department of Signal Theory, Communications and Telematics Engineering, Higher Technical School of Telecommunications Engineering, University of Valladolid, Valladolid, Spain;Department of Signal Theory, Communications and Telematics Engineering, Higher Technical School of Telecommunications Engineering, University of Valladolid, Valladolid, Spain;Department of Signal Theory, Communications and Telematics Engineering, Higher Technical School of Telecommunications Engineering, University of Valladolid, Valladolid, Spain;Department of Signal Theory, Communications and Telematics Engineering, Higher Technical School of Telecommunications Engineering, University of Valladolid, Valladolid, Spain;Department of Signal Theory, Communications and Telematics Engineering, Higher Technical School of Telecommunications Engineering, University of Valladolid, Valladolid, Spain

  • Venue:
  • AIKED'11 Proceedings of the 10th WSEAS international conference on Artificial intelligence, knowledge engineering and data bases
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

The increasing popularity of web map services has motivated the development of more scalable services in the Spatial Data Infrastructures (SDI). Tiled map services have emerged as an scalable alternative to traditional map services. Instead of rendering image maps on the fly, a collection of pre-generated image tiles can be retrieved very fast from a server-side cache. However, storage requirements and start-up time for generating all tiles are often prohibitive for many potentially providers when the cartography covers large areas for multiple rendering scales, which forces to use partial caches containing a subset of the total tiles. This work proposes a descriptive model based on the mining of real-world logs from several nationwide public web map services in Spain. The proposed model is able to determine in advance which areas are likely to be requested in the future based exclusively on past accesses. Tiles that are anticipated to be requested soon can be pre-generated and cached for faster retrieval. As the number of tiles grows exponentially with the rendering resolution level, it is rarely feasible to work with statistics of individual tiles. To overcome this issue, a simplified model is proposed which combines statistics from multiple tiles to reduce the dimension of the tiling space. Simulations demonstrate that significant savings of storage requirements can be achieved by using a partial cache with the proposed model, while maintaining a high cache hit ratio.