Abstract

Many scientific problems can be modeled as computational workflows that integrate data from heterogeneous sources and process such data to derive new results. These data analysis problems are pervasive in the physical and social sciences, as well as in government practice. Therefore, techniques that facilitate the creation of such computational workflows are of critical importance. In our work on the Argos project we are developing techniques to automatically create computational workflows in response to user data requests. As a unifying paradigm, we represent both data access and data processing operations as web services.
We focus on a domain representative of many economic analysis problems: good movement analysis in a large metropolitan area. This domain is of great interest to both social scientists and government practitioners, since understanding the patterns of freight flow is crucial for urban planning and forecasting the economic development of a region. Previous approaches to this type of analysis, mainly surveys, are prohibitively expensive. In Argos we propose to integrate data from secondary sources to estimate the flow of commodities in the Los Angeles metropolitan area, following Gordon and Pan [2001]. The Argos approach presents several advantages. First, it is significantly more cost-effective since the problem is now reduced to automated data integration and processing, instead of labor-intensive surveys. Second, many more questions can be posed and answered, since different data sources and data operations can be composed to derive a variety of novel data for the given domain. Finally, our system …

Metadata

publication

Proceedings of the 2004 annual national conference on Digital government …, 2004

year

2004

publication date

2004/5/24

authors

José Luis Ambite, Genevieve Giuliano, Peter Gordon, Stefan Decker, Andreas Harth, Karanbir Jassar, Qisheng Pan, LanLan Wang

link

https://www.academia.edu/download/97063350/ambite-dgo2004-sd.pdf

resource_link

https://www.academia.edu/download/97063350/ambite-dgo2004-sd.pdf

book

Proceedings of the 2004 annual national conference on Digital government research

pages

1-2