We present a design for the wiki-based classification of concepts within scientific and technical disclosures. Each classification code is a graphic representation of an individual entity together with related entities: the degree of similarity being apparent from what is effectively a one-dimensional ontology. A library of hyper linked graphic representations facilitates the navigation to the correct codes. The same graphic representations form the core of the search engine. A searched concept is defined by a classification code in each of up-to-five facets. Search results are ranked according to the combined similarity of the classification codes in the selected facets. The system may be used to search the prior art or to find novel solutions to problems. We foresee the development of the system to incorporate increasing levels of computer intelligence. We propose the establishment of a global authority, similar to the Wikimedia Foundation, to oversee the development of the system for the benefit of all.