Benchmarking Services

Home

Research

I was involved in the Web-based Benchmark data engine being developed at the Computer Science Department, University of New Orleans.

The objective was to implement a Java and XML Web-based generator that creates a highly flexible benchmark data engine.  This dynamically generated engine should support optimizations of work processes by performance comparisons among companies.  Benchmarking data form a crucial basis for decision and optimization processes.  The final requirements for this engine were:

• Extremely high flexibility, allowing automatic Website generation
• Integration of heterogeneous communication media in this one single system

The high configuration flexibility of such a data engine enables the company to offer new services in a rapid pace without a development project for each service.  The second point concerns the integration of the following communication media: Web-based interaction, direct data exchange, e-mail and fax, to facilitate the usage of benchmarking services for new customers.  The project addresses these points by delivering a Website generator that automatically generates Websites, multiple Web pages based on configuration data in a highly flexible manner.

The highlight of the website generator is its ability to dynamically generate new websites, based on a number of parameters given by the creator of the website. In effect, even users with a non-technical background can create their own websites in minimum amount of time. To start with, this engine was used for the maritime industry, where several shipping companies can create their own websites. They can also upload their data, for example, data pertaining to performance of ships, life & failure of individual parts, mean time before failure of equipment.

Storing this data gives us the chance of analyzing it, which may reveal any interesting patterns in the huge collection of data. A user may wish to compare the services provided by a company with those provided by another company. The search may result in numerical data. For example, the user may want to see the mean time between failures of equipment for a certain ship and may also like to see how the same equipment fared on another ship. This involves a lot of processing. The resulting information can be given to the user in terms of textual data - in the form of numbers, or in a tabular form. This adds an overhead for the user to manually analyze data and understand it. To the human observer, the exploration process can be both time-consuming and even cause him or her to misinterpret the information presented. This combined with a visual data analysis tool forms a powerful benchmarking service.

Benchmarking Engine
Snapshot will be uploaded shortly