Solution #4

Text-mining Based Search Solution

The purpose of this text-mining based solution is to support companies, institutes and inventors for their prior art search in more prompt and accurate ways. Instead of directly inputting complicated search formula, this solution enables users to bring questions with document numbers. By providing search results in the order of similarity among documents, the solution compensate the defect from the existing search solution that was keyword-based.

The major strength of our Smart Trademark Search Solution is to provide a list with not only the strictly identical marks but also the phonetic variants.

Main Features

Automatic extraction of key words within documents

Company's own similarity measuring and allocation algorithm​

Extension of potential synonyms based on LSA (Latent Semantic Analysis) technique

Expansion of cloud computing environment for large volume of document process

Stop-word elimination

Extension of dictionary-based synonyms

Search result clustering

Various different weight value set parameter​