INTELLIGENT MIND-MAP
Intelligent support for informal documents on the tablet PC is hampered by inadequate recognition algorithms and ink reflow techniques. This project addressed both of these problems within the context of mind-maps. Mind-maps are especially attractive as an exemplar of an intelligent ink project as they require a combination of ink support techniques that are applicable to a wide range of other ink document domains. Furthermore the cognitive efficacy of the resulting tool can be evaluated against existing computer-based tools and paper equivalents.
Our prototype tablet PC based mind-mapping tool (Chik, et al, 2007) is designed to support pedagogically-oriented brainstorming. The scenario of use for this project is an educational application of mind-maps: planning a report. Imagine a student planning a report: they start with the main subject in the centre of the mind map and as ideas occur to them add branches and sub-branches, in no particular sequence. When satisfied with the mind-map they convert it into a linear tree structure to commence the detailed report writing. Each branch of the mind map becomes a section and each sub-branch a sub-section.
To provide such an intelligent tool the following functionality is required:
- Basic ink support (drawing, erasing etc) and file persistence.
- Ink grouping and dividing (text/drawing)
- Text and drawing recognition
- Document structural analysis
There is a natural semantic split between drawing and writing ink which humans achieve effortlessly – however it has proved difficult to achieve acceptable results on computers. Identifying ink features to discriminate between drawing and writing strokes is the subject of a current project being undertaken by Blagojevic and Plimmer. Our objective here was to leverage our novel feature recognition based approach and extend it with mind-map-specific features to further improve recognition. These extensions, in addition to permitting us to develop a functional mind-map tool, provide us insight into domain specific specialization of our generic recognition algorithms.
Screenshot of the prototype of an intelligent mind-map software
Publications
Paper Publications
- Chik, V., B. Plimmer, and J. Hosking. (2007). Intelligent Mind-mapping. in OzCHI 2007: Entertaining User Interfaces - Adelaide, Austrailia, 187-190.
Other Publications
- Chik, V. (2007). Intelligent Mind-mapping. MEng Thesis, University of Auckland
This project was partially funded by Microsoft® Research Asia.