Concept maps visualize the understanding and knowledge of users about a given content. Traditionally, concept maps are created manually in dedicated applications. In a master thesis, a web application for semi-automatic creation from a given input text was created . However, in this implementation, concept maps are also created in a dedicated web application, and new text information needs to be explicitly copied into the tool.
The goal of this project is to extend the concept map creation and visualization tool  by integrating it into arbitrary web pages. Existing concepts should be visualized as in-place text highlighting, and relations between concepts should be revealed on demand. It should furthermore be possible to efficiently create new concepts and relations from the textual information on the web page. This way, it should be possible to quickly associate unknown information on web pages with the user’s knowledge expressed in a concept map.
This work should be implemented as an extension to a web browser (e.g., Chrome) so that arbitrary input information can be integrated into the user’s concept maps. The recommended environment consists of a Chrome extension to parse the text content of arbitrary web pages, a Python server to process the content, and a d3.js front-end for the concept map. Alternative frameworks are possible, depending on the preference of the student. Code to generate suggestions for concepts and relations from input text is available from a prior master thesis and can be adapted .