Automatic hotspots recognition and trends prediction in traditional medicine

Exemplified by the field of traditional medicine in this paper, we classify articles extracted from SCIE (2004-2008) medical databases into categories from different perspectives and apply co-word analysis, multidimensional scaling methods to each category. Eventually, two maps are produced for each category, which are the current hotspots map and the trends map. Taking ldquoDrugs, Chinese, Herbalrdquo as an example, its current hotspots map reveals three hot topics: plant extraction, antioxide and cancer cell in 2008. We also find topics concerning cancer in its 2006 trends map, which proved our method regarding high growth rate words as an indicator of the future trends is valid.