Ideagrams: A digital tool for observing ideation processes

Purpose - The paper reports on an ongoing research project of TU Dresden Laboratory of Knowledge Architecture aiming the investigations of the traceability and visualization of upcoming ideas and topics within discussions. Communication and conversation analysis - to explore knowledge processes communication and interaction analyses emerged as a central scientific approach. Hereby knowledge creation and Knowledge Transfer are understood as collective and co-creative effort. Corresponding analysis tools and methods have been developed for the communication- and knowledge creation processes digital media extensively (Faraj et al 2011). However, research focusing on direct and immediate conversation, and not only based on digital media, rarely exists. Development requirements - The existing tools for the analysis of digital communication data are yet not widely applied in the domain of spoken discussions. Whereas communication processes in the digital domain create their data automatically, the data from natural settings have to be extracted laboriously (Tonfoni 2004). Since there are no effective methods on data recording of voice communication yet existing, there are no strong and evident methods on computer aided conversation tracking and analysing existing too. The Ideagram tool tries to overcome this shortcoming. Approach - A prototype of a transcription, visualization and analysis tool was designed, which is able to capture discussions by keywords and analyse them in real-time. The results are presented in several forms: histograms, semantic networks and mixtures between both. These visualizations allow identifying topic and concept dynamics, heuristic paths and creative moments. Central features of the discussion like knowledge communication, orientation for innovation and speech efficiency can be understood and designed. In the different figures - Ideagram - of spoken discussions the logged content is visualised. The program prototype counts the occurrence of the logged words. These can be marked within a chronological re-presentation graph and shows at what time which issue was discussed. By "peaks" and "valley's" it is obviously visible where the talk was most or least active. The prototype tool allows analysing the used phrases according to their frequency and their appearance during the captured conversation. Practical implications & Value - In contrast of conventional protocol and transcription techniques this kind of knowledge mining allows a greater information bandwidth and a more efficient access on core topics, thematic conflicts, idea generation etc. Experiences in very different settings created a very rich data set and allows to state that the application in business and science seems to very useful according to recording, analysing and deepening of spoken discussions. Hence, the Ideagram is still a prototype version and need further investigation and development.