Automatic travelroute detection and generation using heterogeneous geographic data sources and blogs

Beside the traditional sources for travel planning like guidebooks, more and more personalized travelogues, available as blogs or personal websites, are used to gather an extensive basis of decision-making. In this thesis a modular software system was developed, autonomously crawling the web for appropriate travelogues and extracting points of interest from the free text by point of interest identification. The extracted points of interest are enriched with supplementary metadata and joined to routes according to the respective geographical area. These routes are kept by the system in order to be presented to a tourist visiting a certain location.