8.7.1 The past few decades have seen a huge amount of sophisticated code being developed to solve specific, homogenous problems. A person cannot realistically know of all possible solutions to a problem, and expecting a single agent to perform this task is equally unrealistic. To circumvent this resource selection bottleneck, developers have started working on recommender agents. The purpose of a recommender agent is to accept a query from a solver (or mediator) agent about a problem, determine a suitable algorithm that applies to that problem, and finally, direct it to the appropriate location on the net where software implementing the algorithm can be obtained and executed. Such agents are used extensively in commercial search engines and web-based data warehouses.

8.7.2 The organisation of software on the Net and tracking software availability is facilitated by cross-indices of mathematical software such as GAMS. [5] Each recommender agent can provide recommendations for certain class of problems and can also collaborate with other agents to collectively arrive at a recommendation. The PYTHIA agent system [6] provides the recommender agents needed for multidisciplinary simulation. The interface between PYTHIA and the GAMS repository forms the basis of a collaborative software. The PYTHIA agents are based on extensive performance evaluation of GAMS-indexed software. (Drashansky et al., 1999, pg. 51) [top] Notes:

5. For more on GAMS, see: http://gams.nist.gov [back]

6. For more on the PYTHIA agent, see: Weerawana, S. et al. 1996. "PYTHIA: A knowledge-based system to select scientific algorithms." ACM trans. Mathematical Software Volume 22, Number 4. pp 447-468 [back]