期刊名称：International Journal of Electrical and Computer Engineering
电子版ISSN：2088-8708
出版年度：2015
卷号：5
期号：3
页码：483-490
DOI：10.11591/ijece.v5i3.pp483-490
语种：English
出版社：Institute of Advanced Engineering and Science (IAES)
摘要：In this paper we proposed a method which avoids the choice of natural language processing tools such as pos taggers and parsers reduce the processing overhead. Moreover, we suggest a structure to immediately create a large-scale corpus annotated along with disease names, which can be applied to train our probabilistic model. In this proposed work context rank based hierarchical clustering method is applied on different datasets namely colon, Leukemia, MLL medical diseases. Optimal rule filtering algorithm is applied on these datasets to remove unwanted special characters for gene/protein identification. Finally, experimental results show that proposed method outperformed existing methods in terms of time and clusters space.
其他摘要：In this paper we proposed a method which avoids the choice of natural language processing tools such as pos taggers and parsers reduce the processing overhead. Moreover, we suggest a structure to immediately create a large-scale corpus annotated along with disease names, which can be applied to train our probabilistic model. In this proposed work context rank based hierarchical clustering method is applied on different datasets namely colon, Leukemia, MLL medical diseases. Optimal rule filtering algorithm is applied on these datasets to remove unwanted special characters for gene/protein identification. Finally, experimental results show that proposed method outperformed existing methods in terms of time and clusters space.
