Finnish Patent and Registration Office
Many of the duties of the Finnish Patent and Registration Office PRH, such as the granting of patents or the enforcement of copyrights, have to do with detecting similarities. Approving the names of new companies is one of these duties: names that are too close to an existing name are not approved, at least not for companies operating in the same industry. The assessment is based on rules that relate to the uniformities of spelling and phonetic forms of the words. The aim is a company name is unique and distinctive.
Examining existing company names without proper tools was time-consuming, and the task was largely the responsibility of PRH experts. As there was no efficient way for new entrepreneurs to browse existing company names, a rejection for the proposed company name could come as a surprise at a fairly late stage of the start-up process. Lingsoft implemented a Name checking service for companies for PRH to facilitate and speed up the process of checking the registrability of company names.
From a background solution to an open service
In the name checking solution, PRH’s numerous name checking rules and background data were modelled into a self-checking machine leveraging Lingsoft’s technology. The solution was first integrated into PRH’s information system for internal use. In spring 2017, the proven solution was then made into a public service that entrepreneurs could use to compare their idea for a company name against existing company names before submitting the name to PRH for approval.
The uniqueness of Lingsoft’s solution is based on an understanding of the special features of the Finnish language, such as compound words and inflections. The core of the solution is Lingsoft’s smart language analysis. As the machine also takes the phonetic form of the name into account, it can identify bars related to pronunciation as well. The results can be weighted and adjusted by first entering data as training material.
The end result is a solution that returns a complete list of similar names found in the entered data in seconds, saving a considerable amount of time for experts and aspiring entrepreneurs.
