General Approach: . . . Interval Approach: . . . Interval . . . Interval . . . Similar Situation: . . . Towards Secure Let Us Use a Similar . . . How to Represent Sets How to Propagate . . . Cyberinfrastructure for How to Propagate . . . First Example: . . . Sharing Border Information Second Example: . . . How to Compute r ik Distributivity: a · ( b + . . . Ann Gates 1 , Vladik Kreinovich 1 , Luc Longpr´ e 1 , Distributivity: New . . . Paolo Pinheiro da Silva 1 , G. Randy Keller 2 Toy Example with . . . Computation Time What Next? 1 Department of Computer Science Probabilistic Case: In . . . 2 Department of Geological Sciences Acknowledgments University of Texas at El Paso Title Page 500 W. University, El Paso, TX 79968, USA agates@utep.edu, vladik@utep.edu, longpre@utep.edu, ◭◭ ◮◮ paulo@utep.edu, keller@utep.edu ◭ ◮ Page 1 of 17 Go Back Full Screen
General Approach: . . . Interval Approach: . . . 1. Outline Interval . . . Interval . . . • In many border-related issues ranging from economic collaboration to border Similar Situation: . . . security, sharing information is important . Let Us Use a Similar . . . • Sharing is difficult: different countries use different information formats and How to Represent Sets data structures. How to Propagate . . . How to Propagate . . . • Desirable: to facilitate information sharing. First Example: . . . • Our experience: geoinformatics. Second Example: . . . How to Compute r ik • This experience can be applied to border collaboration. Distributivity: a · ( b + . . . • Additional problem: many security-related data are sensitive. Distributivity: New . . . Toy Example with . . . Computation Time What Next? Probabilistic Case: In . . . Acknowledgments Title Page ◭◭ ◮◮ ◭ ◮ Page 2 of 17 Go Back Full Screen
General Approach: . . . Interval Approach: . . . 2. Practical Problem: Need to Combine Geographi- Interval . . . cally Separate Computational Resources Interval . . . Similar Situation: . . . • Problem: Let Us Use a Similar . . . How to Represent Sets – In different domains, there is a large amount of data stored in different How to Propagate . . . locations. How to Propagate . . . – There are many software tools for processing this data, also implemented First Example: . . . at different locations. Second Example: . . . • Users may be interested in different information about this domain. How to Compute r ik Distributivity: a · ( b + . . . – Sometimes, the information required by the user is already stored in one Distributivity: New . . . of the databases . Toy Example with . . . – In other cases, different pieces of the information requested by the user Computation Time are stored at different locations . What Next? – In many other situations, the appropriate answer to the user’s request Probabilistic Case: In . . . requires that we not only collect the relevant data, but that we also use Acknowledgments some data processing algorithms to process this data. Title Page • The need to combine computational resources (data and programs) located ◭◭ ◮◮ at different geographic locations seriously complicates research. ◭ ◮ Page 3 of 17 Go Back Full Screen
General Approach: . . . Interval Approach: . . . 3. Centralization of Computational Resources – Tra- Interval . . . ditional Approach to Combining Computational Re- Interval . . . Similar Situation: . . . sources; Its Advantages and Limitations Let Us Use a Similar . . . How to Represent Sets • Traditional approach: move all the resources to a central location . How to Propagate . . . • Problem: centralization requires a large amount of efforts: How to Propagate . . . First Example: . . . – data are presented in different formats, Second Example: . . . – the existing programs use specific formats, etc. How to Compute r ik Distributivity: a · ( b + . . . • To make the central data depository efficient, it is necessary: Distributivity: New . . . – to reformat all the data, Toy Example with . . . – to rewrite all the data processing programs – so that they become fully Computation Time compatible with the selected formats and with each other, What Next? – etc. Probabilistic Case: In . . . Acknowledgments • Conclusion: the amount of work that is needed for this reformatting and Title Page rewriting is large. ◭◭ ◮◮ • Result: none of these central depositories really succeeded in becoming an easy-to-use centralized database. ◭ ◮ Page 4 of 17 Go Back Full Screen
General Approach: . . . Interval Approach: . . . 4. Cyberinfrastructure – A More Efficient Approach Interval . . . to Combining Computational Resources Interval . . . Similar Situation: . . . • Objective: Let Us Use a Similar . . . How to Represent Sets – provide the users with the efficient way to submit requests How to Propagate . . . – without worrying about the geographic locations of different computa- How to Propagate . . . tional resources First Example: . . . – and avoid centralization with its excessive workloads. Second Example: . . . How to Compute r ik • Main idea: keep all (or at least most) computational resources Distributivity: a · ( b + . . . – at their current locations , Distributivity: New . . . – in their current formats . Toy Example with . . . Computation Time • Specifics: to expedite the use of these resources: What Next? Probabilistic Case: In . . . – we supplement the local computational resources with the “metadata”, i.e., with the information about the formats, algorithms, etc., Acknowledgments Title Page – we “wrap up” the programs and databases with auxiliary programs that provide data compatibility into web services , ◭◭ ◮◮ • General description: we provide a cyberinfrastructure that uses the metadata to automatically combine different computational resources. ◭ ◮ Page 5 of 17 Go Back Full Screen
General Approach: . . . Interval Approach: . . . 5. Cyberinfrastructure: Example Interval . . . Interval . . . • User’s request: a user is interested in using the gravity data to uncover the Similar Situation: . . . geological structure of the Rio Grande region. Let Us Use a Similar . . . • What the system will do: How to Represent Sets How to Propagate . . . – get the gravity data from the UTEP and USGS gravity databases, How to Propagate . . . – convert them to a single format (if necessary), First Example: . . . – forward this data to the program located at San Diego Supercomputer Second Example: . . . Center, and How to Compute r ik Distributivity: a · ( b + . . . – move the results back to the user. Distributivity: New . . . • Comment: this example is exactly what we are designing under the NSF- Toy Example with . . . sponsored Cyberinfrastructure for the Geosciences (GEON) project. Computation Time • General description: this is similar to what other cyberinfrastructure projects What Next? are trying to achieve. Probabilistic Case: In . . . Acknowledgments Title Page ◭◭ ◮◮ ◭ ◮ Page 6 of 17 Go Back Full Screen
General Approach: . . . Interval Approach: . . . 6. What Is Cyberinfrastructure: From the Official NSF Interval . . . Definition Interval . . . Similar Situation: . . . • Source: NSF Blue Ribbon Advisory Panel on Cyberinfrastructure. Let Us Use a Similar . . . How to Represent Sets • Motivation: “a new age has dawned in scientific and engineering research, How to Propagate . . . pushed by continuing progress in How to Propagate . . . – computing, First Example: . . . Second Example: . . . – information, and How to Compute r ik – communication technology, Distributivity: a · ( b + . . . and pulled by the Distributivity: New . . . Toy Example with . . . – expanding complexity, Computation Time – scope, and What Next? – scale Probabilistic Case: In . . . Acknowledgments of today’s challenges.” Title Page • Essence: “The capacity of this technology has crossed thresholds that now make possible a comprehensive ‘cyberinfrastructure’ on which ◭◭ ◮◮ – to build new types of scientific and engineering knowledge environments ◭ ◮ and organizations and – to pursue research in new ways and with increased efficacy.” Page 7 of 17 Go Back Full Screen
General Approach: . . . Interval Approach: . . . 7. What Is Cyberinfrastructure: From the Official NSF Interval . . . Definition (Examples) Interval . . . Similar Situation: . . . • “Such environments and organizations, enabled by cyberinfrastructure, are Let Us Use a Similar . . . increasingly required to address national and global priorities, such as How to Represent Sets How to Propagate . . . – understanding global climate change, How to Propagate . . . – protecting our natural environment, First Example: . . . – applying genomics-proteomics to human health, Second Example: . . . – maintaining national security, How to Compute r ik Distributivity: a · ( b + . . . – mastering the world of nanotechnology, and Distributivity: New . . . – predicting and protecting against natural and human disasters, Toy Example with . . . • as well as to address some of our most fundamental intellectual questions Computation Time such as What Next? Probabilistic Case: In . . . – the formation of the universe and Acknowledgments – the fundamental character of matter.” Title Page ◭◭ ◮◮ ◭ ◮ Page 8 of 17 Go Back Full Screen
Recommend
More recommend
