Taylor Wessing Global Intellectual Property Index 2009

Methodology

The GIPI ratings for different jurisdictions are calculated by a ‘factor assessment model’ built using two distinct sets of input:

Jurisdiction assessments –

To construct the GIPI ratings we use 18,091 assessments drawn from 495 respondents to an online questionnaire.  Respondents assess the competitiveness of the jurisdictions which they know.  The online questionnaire runs continuously to keep the GIPI up-to-date with people’s changing assessments.

Instrumental factors –

Drawn from external sources.  For example, enforcing patents in a jurisdiction is indicated by ‘instrumental factors’ including the number of cases brought, the success rate and the number of patents lawyers.  50 instrumental factors were used in GIPI 2 – these are listed in Appendix C. Not all jurisdictions have data for all instrumental factors and the statistical model takes account of these gaps.

The jurisdiction assessments and instrumental factors are used to build a predictive model of IP jurisdiction competitiveness using support vector machine (SVM) mathematics. SVMs are based upon statistical techniques that classify and model complex historic data in order to make predictions on new data (see below for more on SVMs).

The SVM used for the GIPI provides information about the confidence with which each specific classification is made and the likelihood of other possible classifications.  The predictive model provides the overall index by answering questions such as:

If a London based trade mark lawyer gives France and Germany certain assessments then, based on the instrumental factors for France, Germany and Italy, how would that person assess Italy?

For each of the 16 areas covered by the questionnaire the predictions made by the SVM are combined with the actual assessments to create a matrix of how each respondent assessed or was predicted to have assessed each of the 24 jurisdictions. (In addition to these 24 jurisdictions, respondents were asked to assess Europe as a whole for trade marks (the Community Trade Mark Office), patents (the European Trademark Office) and domains (.eu)). This matrix is used to develop 16 sub-indices, one for each of the areas in the table below:

These sub-indices are the used to create five IP area indices for trade marks, patents, copyright, designs and domains.

The process of creating the GIPI is outlined diagrammatically below:

A full list of the instrumental factors used in the GIPI model is in Appendix D. We have set a number of guidelines to ensure that jurisdiction assessments and instrumental factors are selected and used in a way that will generate a credible, dynamic rating of IP competitiveness.

The guidelines for independent indices used as instrumental factors are:

indices should come from a reputable body and be derived by a sound methodology;

indices should be readily available;

relevant indices can be added to the GIPI model at any time;

updates to the indices are collected and collated every six months;

no weightings need be applied to the indices - the SVM calculates which factors are the most influential;

indices are be entered into the GIPI model as directly as possible, whether they are a rank, a derived score, a value, a distribution around a mean or a distribution around a benchmark;

if a factor is at a national level, the score is used for all areas in that country – nation based factors are avoided if area or city based factors are available;

if an index has multiple values for a jurisdiction, the most relevant value is used (and the method for judging relevance noted);

if an index does not contain a value for a particular jurisdiction, a blank is entered against that jurisdiction (no average or mean is used). Only indices which have values for at least half the GIPI jurisdictions are used.

Creating the Global IP Index does not involve totalling or averaging instrumental factors. An approach involving totalling and averaging would involve a number of difficulties:

indices are published in a variety of different forms: an average or base point of 100 with scores above and below this; a simple ranking; actual values; a composite ‘score‘;

indices would have to be normalised, e.g. in some indices a high score is positive while in others a low score is positive;

not all jurisdictions are included in all indices;

the indices would have to be weighted.

Back to top

© 2009 | Taylor Wessing LLP
Terms of use | Privacy policy | Regulatory | Environmental