Languages

General D-Sight Methodology

Enrich Data

D-Sight offers you the possibility to enrich the data of your multi-criteria problem by taking into account how you want to evaluate the alternatives for each criterion.

Working with raw data is rarely meaningful and is, most of the time, not efficient.

The multi-criteria evaluation/prioritization/decision problem can always be written in a “decision matrix” wich is illustrated below. In each line you find an alternative and the criteria are represented in the column.

  c1 c2 ... ... ck
a1 152357 yes ... ... 100
a2 216375 no ... ... 32
... ... ... ... ... ...
... ... ... ... ... ...
an 198542 yes ... ... 73

What D-Sight allows you to do, before making any decision analysis, is to enrich this first decision matrix. This can be done by making preference modeling. D-Sight offers two methods to do it:

Meaningful Results

These methods give you the possibility to easily obtain a score matrix:

  score for c1 score for c2 ... ... score for ck
a1 10,0 10 ... ... 7,8
a2 03,7 0 ... ... 6,2
... ... ... ... ... ...
... ... ... ... ... ...
an 06,8 10 ... ... 6,9

Giving a weight to each criterion allows computing a global score for each alternative. The difference with the weighted sum is that you do not sum the raw data. Instead the enriched data is summed. That means that the importance of the criteria is actually taken into account when the global score is computed. The global score is therefore meaningful and represents the real value of the alternative.