Integrating the World: The WorldInfo Assistant Craig A. Knoblock †‡ , Jose Luis Ambite † , Steven Minton ‡ , Cyrus Shahabi † , Mohammad Kolahdouzan † , Maria Muslea † , Jean Oh † , and Snehal Thakkar † † University of Southern California ‡ Fetch Technologies 4676 Admiralty Way, Suite 1001 4676 Admiralty Way, 10th floor Marina del Rey, CA 90292 Marina del Rey, CA 90292 Abstract The Internet is an extraordinary re- source for information about countries throughout the world. There is detailed information about weather, geography, transportation, politics, news, etc. This data comes in a variety of forms in- cluding web pages, databases, maps, satellite im- agery, newspapers, online radio and television, and so on. The problem, of course, is how to navi- gate through this abundance of information with- out getting bogged down in the details of the loca- tion, organization, and format of the information. Towards this end we have developed the WorldInfo Assistant, which extracts and integrates geographic- related data about countries throughout the world. Keywords: Information integration, wrappers, constraint reasoning, Heracles, WorldInfo As- sistant 1 Introduction The Internet provides an enormous amount of data about countries throughout the world, but the problem is how to locate, organize and present all of this information in a way that is natural and simple to use. To address this chal- lenge, we have developed an application called the WorldInfo Assistant, which integrates all of this geographical and multimedia informa- tion about countries into a single easy-to-use framework. The user specifies the region of the world and the timeframe of interest, and the system provides the information relative to these parameters organized into a hierarchical representation that can be conveniently navi- gated. There are two key underlying technologies that make the WorldInfo Assistant possible. First, we have developed a general integra- tion framework, called Heracles, for building information assistants. The core of Heracles is an interactive constraint reasoning system that makes it easy for the user to navigate through the large amount of available data without be- ing overwhelmed with information. Second, we have developed a set of tools for building wrap- pers that can turn semistructured data sources into structured sources. This allows the sys- tem to access and integrate data from online Web sources since much of the information is available in this form and is not available as databases. 2 The WorldInfo Assistant The WorldInfo Assistant provides integrated access to a variety of sources about countries throughout the world. The initial system pro- vides only a small fraction of the many sources available, but the system is very extensible and demonstrates the ability to integrate and orga- nize a wide variety of sources. In this section we describe the key components of the system. 2.1 The Heracles Constraint Rea- soning System We have developed a hierarchical constraint reasoning system called Heracles [3], which provides the underlying reasoning support for