Mining Ideas from Textual Information
(Technological Idea Miner)

Fraunhofer INT

Text mining refers generally to the process of deriving high quality information from unstructured texts. Unstructured texts come in many shapes and sizes. It may be stored in research papers, articles in technical periodicals, reports, documents, web pages etc. Here we introduce a new approach for finding new ideas and inventions in unstructured technological texts.

This text mining approach follows the statements of technique philosophy. Therefore a new idea, invention or innovation represents not only a new mean, but a new purpose and mean combination. By systematic analysis of the purpose and mean combinations in new texts compared to specialized reference collections, an automatic collection of new ideas and inventions can be realized.

Here we demonstrate our first application based on the idea mining approach.

Technological Idea Miner

The idea mining approach is published in Expert Systems with Applications. For evaluation, we use patent data. You can find them here. Additionally, we use a Jaccard's Coefficient application for comparing our Technological Idea Miner to the baseline. You find this application here.

Screenshots

Fraunhofer-Gesellschaft
Fraunhofer-Gesellschaft