Conceptual Clustering of Patents

Enhancing the insight for the decision making process in Strategic Technology Planning

Kevin Kruijthoff

This webpage is accompanying my thesis on the 'Conceptual Clustering of Patents' and is aimed to enable its readers to interact with the visualizations described, as well as to download the prototype software. Before exploring the visualizations I recommend to watch the short embedded video. Which contains an overview of the tool, an explanation on the currently included possibilities and suggested additions to further enhance the possibilities of the decision support tool.

Visualization Notes:
  1. Be aware that all visualizations except the interactive demo are large in filesize and thus may require some time to parse its elements.
  2. There is a limited support of the SVG .html element in many browsers. To make sure you're also able to see the node explanations in the visualizations please use the Google Chrome browser when exploring the visualizations.
* This visualization is created for comperative purposes, for which the indexation of the clustering input is created using the text-mining python script as presented in: "Cunningham, S. W. and J. H. Kwakkel (2015), Analytics and Tech Mining for Engineering Managers, Momentum Press: New York."
The developed prototype entails a three step process. First, the patent data is retrieved from a database and indexed by the Python text-mining software, to be stored in a JSON file. Secondly, the indexed patents are transformed into a hierarchical relational structure using the Python conceptual clustering software and stored in a JSON file. Finally, the hierarchical relational structure is visualized in an online environment which is created using the d3.js javascript library.