Browse Prior Art Database

System and Device of Evaluation Farm Based on Multiple Criteria Decision Making Approach and Knowledge Base

IP.com Disclosure Number: IPCOM000031875D
Original Publication Date: 2004-Oct-15
Included in the Prior Art Database: 2004-Oct-15
Document File: 6 page(s) / 251K

Publishing Venue

IBM

Abstract

The disclosed system and device use Multiple Attribute Decision Making (MADM) approach, and provide evaluation result for the input alternatives by providing their rank or their evaluation scores. The system can plug evaluation models, evaluation algorithms by standard interfaces, and provide standard interfaces for other systems to access evaluation results. An evaluation task is used to describe the evaluation object, relationship model and evaluation algorithm.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 23% of the total text.

Page 1 of 6

System and Device of Evaluation Farm Based on Multiple Criteria Decision Making Approach and Knowledge Base

1. Background: What is the problem solved by your invention? Describe known solutions to this problem (if any). What are the drawbacks of such known solutions, or why is an additional solution required? Cite any relevant technical documents or references.

Evaluation is very important for an enterprise since most of decision is based on evaluation. Many activities, from business level to IT level, such as partner selection, employee performance evaluation, the IT system evaluation, are related to evaluation. Some evaluation is of lightweight, such as selecting a stationery vendor from three candidates, maybe two factors, e.g. price and quality, are enough for the partner evaluation. However, some evaluation is heavy weighted; some scenarios of these kinds of evaluation are shown below.

Scenario 1 (Business level): currently, the telecom operators in Asia have a lot of content providers as their partners. These content providers provide subscribing content services through the telecom network. The operators share the revenue from network usage with the content providers. So their partner relationship is "win-win". However, some content providers get out of line and dissatisfy the subscribers, which effects operators' reputation and even make operators lose subscribers and revenue. Therefore, the partner evaluation is very important for operators because it can help operators identify the value of their partners, such that which partner is valuable for them and which one should be given up. This kind of evaluation is ongoing for ever, and it needs to receive data from operator's existing systems, and output evaluation result to operator's other systems, such as partner encouragement and punishment system, gateway. When the requirements change, the evaluation model and task should be changed on demand. This kind of evaluation requires a flexible framework which can easily integrate with other existing systems.

Scenario 2 (Business level): there is an evaluation agent that provides evaluation service to customers. The evaluation tasks from customers are quite different, some tasks are similar to the scenario 1, and some are different. The customer may require that the evaluation models are authoritative, cross, and selectable. Even some customers demand the evaluation model can be adjusted according to their specific requirements. All these requirements call for a flexible framework of evaluation system.

Scenario 3 (IT level): a composite service consists of many other services. For a specific service, there will be many candidate suppliers. We hope that each candidate we selected for a certain service in a composite service should be the best supplier. However, how to identify the best candidate? A real time evaluation system can evaluate the candidates based on both time and event and help the composite service select service c...