An optimized bibliometric method was applied in this work to evaluate global scientific production of data mining papers of the Science Citation Index (SCI). In compared with traditional bibliometric keyword analysis, a semantic words class was established by applying the text extraction mode to remove noise in the abstract and combining with the core relative phrases retrieved from keywords to get the sample for further experiment. The analysis shows a high correlation between title and keywords, and the title reports less information than keywords does. Also, keywords provide more positive guidance to know and be familiar with the status and trends of this field. In addition, there are distinctions among those semantic words used in publications from the ten most productive countries in data mining research. Generally speaking, the research results can be extended to investigate the roadmap for future research, and this innovative propose is provided with instructive meaning for valuable information retrieval.
Dai, Lu; Ding, Lixin; Lei, Yunwen; and Tian, Yangge
"A study of data mining trend through the optimized bibliometric methodology based on SCI database from 1993 to 2011,"
Applied Mathematics & Information Sciences: Vol. 06:
3, Article 42.
Available at: https://digitalcommons.aaru.edu.jo/amis/vol06/iss3/42