A novel method for visualization of a fuzzy or crisp topic set is developed. The method maps the set's topics to higher ranks of the taxonomy tree of the field. The method involves a penalty function summing penalties for the chosen "head subjects" together with penalties for emerging "gaps" and "offshoots". The method finds a mapping minimizing the penalty function in recursive steps involving two different scenarios, that of ‘gaining a head subject' and that of ‘not gaining a head subject'. We illustrate the method by applying it to illustrative and real-world data.
|Title of host publication||LNCS|
|Publication status||Published - 1 Jan 2011|
|Event||Procs. of the 4th International Conference on Pattern Recognition and Machine Intelligence (PReMI) - |
Duration: 1 Jan 2011 → …
|Conference||Procs. of the 4th International Conference on Pattern Recognition and Machine Intelligence (PReMI)|
|Period||1/01/11 → …|