International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
|
Volume 70 - Issue 20 |
Published: May 2013 |
Authors: Anirban Chakrabarty |
![]() |
Anirban Chakrabarty . A Framework for Medical Text Mining using a Novel Categorical Clustering Algorithm. International Journal of Computer Applications. 70, 20 (May 2013), 19-25. DOI=10.5120/12184-8240
@article{ 10.5120/12184-8240, author = { Anirban Chakrabarty }, title = { A Framework for Medical Text Mining using a Novel Categorical Clustering Algorithm }, journal = { International Journal of Computer Applications }, year = { 2013 }, volume = { 70 }, number = { 20 }, pages = { 19-25 }, doi = { 10.5120/12184-8240 }, publisher = { Foundation of Computer Science (FCS), NY, USA } }
%0 Journal Article %D 2013 %A Anirban Chakrabarty %T A Framework for Medical Text Mining using a Novel Categorical Clustering Algorithm%T %J International Journal of Computer Applications %V 70 %N 20 %P 19-25 %R 10.5120/12184-8240 %I Foundation of Computer Science (FCS), NY, USA
The fast growth of medical records provides new opening for meaningful information retrieval in clinical diagnosis and treatment. Although nursing and pathology records provide a complete account of patient's information they are not fully utilized while taking major decisions of surgery or chemo therapy on patients. This research proposes a Minimum spanning tree algorithm to develop k-clusters of training data related to different liver diseases which are validated using Silhouette coefficient. A text classification algorithm is developed using cluster centers as training samples which uses a similarity measure to classify the categorical data. Simulation results show that the algorithm proposed can lower the calculation complexity and improve the accuracy of established text classification algorithms like k-NN. This research can serve as a medical diagnosis tool for classifying patient records and reveal important vocabularies that characterize nursing and pathology records.