In the final three articles in my series looking at algorithms for measuring the similarity of marks¹, I extend the ideas to provide a more sophisticated and adaptable framework.

In article #4 (‘Further developing colour mark similarity measurement - Part II: Defining a similarity score’), the concept of the ‘distance’ between two colours in RGB space is utilised to define a similarity score between any pair of colours. The formulation of a score is better aligned with the familiar trademark assessment framework involving degrees of similarity, but provides the potential for a more quantitative and continuous approach (and is consistent with the scoring approach proposed for use with word marks).

In article #5 (‘Further developing word mark similarity measurement - Part II: Defining an improved similarity score’), the previous algorithm proposed for quantifying the overall visual (spelling) and aural (pronunciation) similarity between word marks is modified, by using a superior IPA (International Phonetic Alphabet) representation of the marks to determine the phonetic (aural) similarity. The updated algorithm performs (subjectively) well in ranking pairs of marks, and offers a repeatable and consistent framework, with applications not only in the obvious realm of trademark disputes, but also in areas such as the possibility of prioritising the outputs from trademark watching services.

The final article (#6) in the series (‘Further developing colour mark similarity measurement - Part III: A method for sorting colours’) looks at an ‘add-on’ for the colour analysis framework; namely the ability to sort colours into a convenient order based on the position of their dominant or ‘base’ shade in the colour spectrum. The associated algorithm has applications not only in reviewing colour marks and assessing potential ‘clashes’, but also in allowing a more informed approach for potentially formulating guidelines regarding the ‘thresholds’ up to which it might be appropriate for colour-mark protection to apply.