A Distributed K-Nearest-Neighbor Algorithm For Text Categorization
Authors Name:
SUMAN SAHU
, DR.ABHA CHOUBEY
Author Reg. ID:
IJNRD_170133
Published Paper Id:
IJNRD1709004
Published In:
Volume 2 Issue 9, September-2017
Abstract:
Text categorization is the application of text mining. Content classification is a supervised learning technique, it plays important role for indexing of document like different applications. Content order has abundant applications, in several fields and for different sorts of information. Numerous issues identified with information stockpiling, administration and recovery can be defined as far as content order. Clustering plays vital part in text mining. K-means clustering is widely used text categorization technique, still more work can be carried out to improve the performance of k-means text classification technique. In this paper we have proposed parallelization of the renowned k-means clustering algorithm. The parallel implementation of k-means uses data parallelism. In this paper we have compared the performance of parallel k-means text classification with sequential k-means with respect to time factor i.e. total time required for content classification and eventually we have calculated the F-measure value by calculating precision and recall which decides what percentage of messages were classified correctly.
Keywords:
KNN;Recall; F-measure; Precision
Cite Article:
"A Distributed K-Nearest-Neighbor Algorithm For Text Categorization", International Journal of Novel Research and Development (www.ijnrd.org), ISSN:2456-4184, Vol.2, Issue 9, page no.10-13, September-2017, Available :http://www.ijnrd.org/papers/IJNRD1709004.pdf
Facebook Twitter Google+ Pinterest LinkedIn