Each dimension could be analysed on its own. Assigning a high or a low value of relevance based on analysis of just a single factor may seem justifyable but somewhat arbitrary. A major benchmark for each relevance value must be the proportion in respect to each other relevance value. Relevance has to be judged analysing all dimensions simultaneously. That leads to the notion of representing relevance hypothesis by means of vectors. Vectors are just another way of representing the relevance of all dimensions simultaneously.